d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
dcceabd344
freebsd-dev/sys/dev/fb/creator.c

1129 lines
30 KiB
C
Raw Normal View History

Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
/*-
* Copyright (c) 2003 Jake Burkholder.
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
* Copyright (c) 2005 - 2006 Marius Strobl <marius@FreeBSD.org>
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/bus.h>
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#include <sys/conf.h>
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
#include <sys/consio.h>
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
#include <sys/fbio.h>
- Add missing <sys/module.h>. [1] - Remove unused includes. - Sort includes. Reported by: Pyun YongHyeon <yongari@kt-is.co.kr> [1]
2004-07-09 23:12:22 +00:00
#include <sys/kernel.h>
#include <sys/module.h>
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#include <sys/resource.h>
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
#include <dev/ofw/ofw_bus.h>
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#include <dev/ofw/openfirm.h>
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
#include <machine/bus.h>
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#include <machine/bus_private.h>
#include <machine/ofw_machdep.h>
#include <machine/resource.h>
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
#include <machine/sc_machdep.h>
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#include <sys/rman.h>
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
#include <dev/fb/fbreg.h>
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#include <dev/fb/creatorreg.h>
Move the gallant 12 x 22 font data from a .h to a .c so it doesn't need to be compiled into every driver making use of it. Use a const instance of struct gfb_font for this as the font isn't intended to be changed at run-time and in order to accompany the font data with height and width info.
2007-06-16 21:48:50 +00:00
#include <dev/fb/gfb.h>
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
#include <dev/syscons/syscons.h>
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#define CREATOR_DRIVER_NAME "creator"
struct creator_softc {
video_adapter_t sc_va; /* XXX must be first */
phandle_t sc_node;
struct cdev *sc_si;
struct resource *sc_reg[FFB_NREG];
bus_space_tag_t sc_bt[FFB_NREG];
bus_space_handle_t sc_bh[FFB_NREG];
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
u_long sc_reg_size;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
u_int sc_height;
u_int sc_width;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
u_int sc_xmargin;
u_int sc_ymargin;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
Move the gallant 12 x 22 font data from a .h to a .c so it doesn't need to be compiled into every driver making use of it. Use a const instance of struct gfb_font for this as the font isn't intended to be changed at run-time and in order to accompany the font data with height and width info.
2007-06-16 21:48:50 +00:00
const u_char *sc_font;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
int sc_bg_cache;
int sc_fg_cache;
int sc_fifo_cache;
int sc_fontinc_cache;
int sc_fontw_cache;
int sc_pmask_cache;
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
u_int sc_flags;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
#define CREATOR_AFB (1 << 0)
#define CREATOR_CONSOLE (1 << 1)
#define CREATOR_CUREN (1 << 2)
#define CREATOR_CURINV (1 << 3)
#define CREATOR_PAC1 (1 << 4)
};
#define FFB_READ(sc, reg, off) \
bus_space_read_4((sc)->sc_bt[(reg)], (sc)->sc_bh[(reg)], (off))
#define FFB_WRITE(sc, reg, off, val) \
bus_space_write_4((sc)->sc_bt[(reg)], (sc)->sc_bh[(reg)], (off), (val))
#define C(r, g, b) ((b << 16) | (g << 8) | (r))
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
static const uint32_t const creator_cmap[] = {
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
C(0x00, 0x00, 0x00), /* black */
C(0x00, 0x00, 0xff), /* blue */
C(0x00, 0xff, 0x00), /* green */
C(0x00, 0xc0, 0xc0), /* cyan */
C(0xff, 0x00, 0x00), /* red */
C(0xc0, 0x00, 0xc0), /* magenta */
C(0xc0, 0xc0, 0x00), /* brown */
C(0xc0, 0xc0, 0xc0), /* light grey */
C(0x80, 0x80, 0x80), /* dark grey */
C(0x80, 0x80, 0xff), /* light blue */
C(0x80, 0xff, 0x80), /* light green */
C(0x80, 0xff, 0xff), /* light cyan */
C(0xff, 0x80, 0x80), /* light red */
C(0xff, 0x80, 0xff), /* light magenta */
C(0xff, 0xff, 0x80), /* yellow */
C(0xff, 0xff, 0xff), /* white */
};
#undef C
static const struct {
vm_offset_t virt;
vm_paddr_t phys;
vm_size_t size;
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
} const creator_fb_map[] = {
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
{ FFB_VIRT_SFB8R, FFB_PHYS_SFB8R, FFB_SIZE_SFB8R },
{ FFB_VIRT_SFB8G, FFB_PHYS_SFB8G, FFB_SIZE_SFB8G },
{ FFB_VIRT_SFB8B, FFB_PHYS_SFB8B, FFB_SIZE_SFB8B },
{ FFB_VIRT_SFB8X, FFB_PHYS_SFB8X, FFB_SIZE_SFB8X },
{ FFB_VIRT_SFB32, FFB_PHYS_SFB32, FFB_SIZE_SFB32 },
{ FFB_VIRT_SFB64, FFB_PHYS_SFB64, FFB_SIZE_SFB64 },
{ FFB_VIRT_FBC, FFB_PHYS_FBC, FFB_SIZE_FBC },
{ FFB_VIRT_FBC_BM, FFB_PHYS_FBC_BM, FFB_SIZE_FBC_BM },
{ FFB_VIRT_DFB8R, FFB_PHYS_DFB8R, FFB_SIZE_DFB8R },
{ FFB_VIRT_DFB8G, FFB_PHYS_DFB8G, FFB_SIZE_DFB8G },
{ FFB_VIRT_DFB8B, FFB_PHYS_DFB8B, FFB_SIZE_DFB8B },
{ FFB_VIRT_DFB8X, FFB_PHYS_DFB8X, FFB_SIZE_DFB8X },
{ FFB_VIRT_DFB24, FFB_PHYS_DFB24, FFB_SIZE_DFB24 },
{ FFB_VIRT_DFB32, FFB_PHYS_DFB32, FFB_SIZE_DFB32 },
{ FFB_VIRT_DFB422A, FFB_PHYS_DFB422A, FFB_SIZE_DFB422A },
{ FFB_VIRT_DFB422AD, FFB_PHYS_DFB422AD, FFB_SIZE_DFB422AD },
{ FFB_VIRT_DFB24B, FFB_PHYS_DFB24B, FFB_SIZE_DFB24B },
{ FFB_VIRT_DFB422B, FFB_PHYS_DFB422B, FFB_SIZE_DFB422B },
{ FFB_VIRT_DFB422BD, FFB_PHYS_DFB422BD, FFB_SIZE_DFB422BD },
{ FFB_VIRT_SFB16Z, FFB_PHYS_SFB16Z, FFB_SIZE_SFB16Z },
{ FFB_VIRT_SFB8Z, FFB_PHYS_SFB8Z, FFB_SIZE_SFB8Z },
{ FFB_VIRT_SFB422, FFB_PHYS_SFB422, FFB_SIZE_SFB422 },
{ FFB_VIRT_SFB422D, FFB_PHYS_SFB422D, FFB_SIZE_SFB422D },
{ FFB_VIRT_FBC_KREG, FFB_PHYS_FBC_KREG, FFB_SIZE_FBC_KREG },
{ FFB_VIRT_DAC, FFB_PHYS_DAC, FFB_SIZE_DAC },
{ FFB_VIRT_PROM, FFB_PHYS_PROM, FFB_SIZE_PROM },
{ FFB_VIRT_EXP, FFB_PHYS_EXP, FFB_SIZE_EXP },
};
#define CREATOR_FB_MAP_SIZE \
(sizeof(creator_fb_map) / sizeof(creator_fb_map[0]))
Move the gallant 12 x 22 font data from a .h to a .c so it doesn't need to be compiled into every driver making use of it. Use a const instance of struct gfb_font for this as the font isn't intended to be changed at run-time and in order to accompany the font data with height and width info.
2007-06-16 21:48:50 +00:00
extern const struct gfb_font gallant12x22;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
static struct creator_softc creator_softc;
static struct bus_space_tag creator_bst_store[FFB_FBC];
static device_probe_t creator_bus_probe;
static device_attach_t creator_bus_attach;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
static device_method_t creator_bus_methods[] = {
DEVMETHOD(device_probe, creator_bus_probe),
DEVMETHOD(device_attach, creator_bus_attach),
{ 0, 0 }
};
static devclass_t creator_devclass;
DEFINE_CLASS_0(creator, creator_bus_driver, creator_bus_methods,
sizeof(struct creator_softc));
DRIVER_MODULE(creator, nexus, creator_bus_driver, creator_devclass, 0, 0);
DRIVER_MODULE(creator, upa, creator_bus_driver, creator_devclass, 0, 0);
static d_open_t creator_fb_open;
static d_close_t creator_fb_close;
static d_ioctl_t creator_fb_ioctl;
static d_mmap_t creator_fb_mmap;
static struct cdevsw creator_fb_devsw = {
.d_version = D_VERSION,
.d_flags = D_NEEDGIANT,
.d_open = creator_fb_open,
.d_close = creator_fb_close,
.d_ioctl = creator_fb_ioctl,
.d_mmap = creator_fb_mmap,
.d_name = "fb",
};
static void creator_cursor_enable(struct creator_softc *sc, int onoff);
static void creator_cursor_install(struct creator_softc *sc);
static void creator_shutdown(void *xsc);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
static int creator_configure(int flags);
static vi_probe_t creator_probe;
static vi_init_t creator_init;
static vi_get_info_t creator_get_info;
static vi_query_mode_t creator_query_mode;
static vi_set_mode_t creator_set_mode;
static vi_save_font_t creator_save_font;
static vi_load_font_t creator_load_font;
static vi_show_font_t creator_show_font;
static vi_save_palette_t creator_save_palette;
static vi_load_palette_t creator_load_palette;
static vi_set_border_t creator_set_border;
static vi_save_state_t creator_save_state;
static vi_load_state_t creator_load_state;
static vi_set_win_org_t creator_set_win_org;
static vi_read_hw_cursor_t creator_read_hw_cursor;
static vi_set_hw_cursor_t creator_set_hw_cursor;
static vi_set_hw_cursor_shape_t creator_set_hw_cursor_shape;
static vi_blank_display_t creator_blank_display;
static vi_mmap_t creator_mmap;
static vi_ioctl_t creator_ioctl;
static vi_clear_t creator_clear;
static vi_fill_rect_t creator_fill_rect;
static vi_bitblt_t creator_bitblt;
static vi_diag_t creator_diag;
static vi_save_cursor_palette_t creator_save_cursor_palette;
static vi_load_cursor_palette_t creator_load_cursor_palette;
static vi_copy_t creator_copy;
static vi_putp_t creator_putp;
static vi_putc_t creator_putc;
static vi_puts_t creator_puts;
static vi_putm_t creator_putm;
static video_switch_t creatorvidsw = {
.probe = creator_probe,
.init = creator_init,
.get_info = creator_get_info,
.query_mode = creator_query_mode,
.set_mode = creator_set_mode,
.save_font = creator_save_font,
.load_font = creator_load_font,
.show_font = creator_show_font,
.save_palette = creator_save_palette,
.load_palette = creator_load_palette,
.set_border = creator_set_border,
.save_state = creator_save_state,
.load_state = creator_load_state,
.set_win_org = creator_set_win_org,
.read_hw_cursor = creator_read_hw_cursor,
.set_hw_cursor = creator_set_hw_cursor,
.set_hw_cursor_shape = creator_set_hw_cursor_shape,
.blank_display = creator_blank_display,
.mmap = creator_mmap,
.ioctl = creator_ioctl,
.clear = creator_clear,
.fill_rect = creator_fill_rect,
.bitblt = creator_bitblt,
.diag = creator_diag,
.save_cursor_palette = creator_save_cursor_palette,
.load_cursor_palette = creator_load_cursor_palette,
.copy = creator_copy,
.putp = creator_putp,
.putc = creator_putc,
.puts = creator_puts,
.putm = creator_putm
};
VIDEO_DRIVER(creator, creatorvidsw, creator_configure);
extern sc_rndr_sw_t txtrndrsw;
RENDERER(creator, 0, txtrndrsw, gfb_set);
RENDERER_MODULE(creator, gfb_set);
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
static const u_char const creator_mouse_pointer[64][8] __aligned(8) = {
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
{ 0x00, 0x00, }, /* ............ */
{ 0x80, 0x00, }, /* *........... */
{ 0xc0, 0x00, }, /* **.......... */
{ 0xe0, 0x00, }, /* ***......... */
{ 0xf0, 0x00, }, /* ****........ */
{ 0xf8, 0x00, }, /* *****....... */
{ 0xfc, 0x00, }, /* ******...... */
{ 0xfe, 0x00, }, /* *******..... */
{ 0xff, 0x00, }, /* ********.... */
{ 0xff, 0x80, }, /* *********... */
{ 0xfc, 0xc0, }, /* ******..**.. */
{ 0xdc, 0x00, }, /* **.***...... */
{ 0x8e, 0x00, }, /* *...***..... */
{ 0x0e, 0x00, }, /* ....***..... */
{ 0x07, 0x00, }, /* .....***.... */
{ 0x04, 0x00, }, /* .....*...... */
{ 0x00, 0x00, }, /* ............ */
{ 0x00, 0x00, }, /* ............ */
{ 0x00, 0x00, }, /* ............ */
{ 0x00, 0x00, }, /* ............ */
{ 0x00, 0x00, }, /* ............ */
{ 0x00, 0x00, }, /* ............ */
};
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
static inline void creator_ras_fifo_wait(struct creator_softc *sc, int n);
static inline void creator_ras_setfontinc(struct creator_softc *sc, int fontinc);
static inline void creator_ras_setfontw(struct creator_softc *sc, int fontw);
static inline void creator_ras_setbg(struct creator_softc *sc, int bg);
static inline void creator_ras_setfg(struct creator_softc *sc, int fg);
static inline void creator_ras_setpmask(struct creator_softc *sc, int pmask);
static inline void creator_ras_wait(struct creator_softc *sc);
static inline void
creator_ras_wait(struct creator_softc *sc)
{
int ucsr;
int r;
for (;;) {
ucsr = FFB_READ(sc, FFB_FBC, FFB_FBC_UCSR);
if ((ucsr & (FBC_UCSR_FB_BUSY | FBC_UCSR_RP_BUSY)) == 0)
break;
r = ucsr & (FBC_UCSR_READ_ERR | FBC_UCSR_FIFO_OVFL);
if (r != 0)
FFB_WRITE(sc, FFB_FBC, FFB_FBC_UCSR, r);
}
}
static inline void
creator_ras_fifo_wait(struct creator_softc *sc, int n)
{
int cache;
cache = sc->sc_fifo_cache;
while (cache < n)
cache = (FFB_READ(sc, FFB_FBC, FFB_FBC_UCSR) &
FBC_UCSR_FIFO_MASK) - 8;
sc->sc_fifo_cache = cache - n;
}
static inline void
creator_ras_setfontinc(struct creator_softc *sc, int fontinc)
{
if (fontinc == sc->sc_fontinc_cache)
return;
sc->sc_fontinc_cache = fontinc;
creator_ras_fifo_wait(sc, 1);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_FONTINC, fontinc);
creator_ras_wait(sc);
}
static inline void
creator_ras_setfontw(struct creator_softc *sc, int fontw)
{
if (fontw == sc->sc_fontw_cache)
return;
sc->sc_fontw_cache = fontw;
creator_ras_fifo_wait(sc, 1);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_FONTW, fontw);
creator_ras_wait(sc);
}
static inline void
creator_ras_setbg(struct creator_softc *sc, int bg)
{
if (bg == sc->sc_bg_cache)
return;
sc->sc_bg_cache = bg;
creator_ras_fifo_wait(sc, 1);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_BG, bg);
creator_ras_wait(sc);
}
static inline void
creator_ras_setfg(struct creator_softc *sc, int fg)
{
if (fg == sc->sc_fg_cache)
return;
sc->sc_fg_cache = fg;
creator_ras_fifo_wait(sc, 1);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_FG, fg);
creator_ras_wait(sc);
}
static inline void
creator_ras_setpmask(struct creator_softc *sc, int pmask)
{
if (pmask == sc->sc_pmask_cache)
return;
sc->sc_pmask_cache = pmask;
creator_ras_fifo_wait(sc, 1);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_PMASK, pmask);
creator_ras_wait(sc);
}
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
/*
* video driver interface
*/
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
static int
creator_configure(int flags)
{
struct creator_softc *sc;
phandle_t chosen;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
phandle_t output;
- Use an ihandle_t to store the stdout instance handle instead of a phandle_t. Since both are typedefed to unsigned int, this is more or less cosmetic. - Fix the code that determines whether a creator instance was used for firmware output (and should not be blanked on initialization). Since r1.2 of dev/fb/creator.c, this consisted comparing a handle of an instance of a package with a handle of the package itself. Use the test from r1.1, which utilizes OF_instance_to_package(). Submitted by: Marius Strobl <marius@alchemy.franken.de>
2004-04-04 12:52:22 +00:00
ihandle_t stdout;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
bus_addr_t addr;
char buf[sizeof("SUNW,ffb")];
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
int i;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
int space;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
- In creator_configure() when probed for the high-level console return the number of registered adapters instead of determining again whether stdout is a supported card (and which might have failed to attach and register). - Drop creator_set_mode() and move the relevant parts to creator_fill_rect() and creator_putc() respectively. This is a bit cleaner than having to make sure that creator_set_mode() was called before creator_fill_rect() or creator_putc() are used and matches better what Xorg does. - Fix a bug in the handling of the FBIOSCURSOR IOCTL; the code was meant to return ENODEV for all invocations expect when used to disable the cursor and not just when used for enabling the cursor. - In case the adapter is the OFW stdout move its OFW cursor to the start of the last line on halt so OFW output doesn't get intermixed with what FreeBSD left on the screen. With hindsight this is what the faking of a hardware cursor which was removed in the last revision really was about, i.e. to keep the OFW updated about the current cursor position. The new approach however is simpler while producing the same result and doesn't cause the first letter of the OFW output to be turned into a blank and a newline. - Add variable names to the prototypes of creator_cursor_*() which were added in the last revision and list them alphabetically in order to match the style of this file.
2005-06-04 21:15:27 +00:00
/*
* For the high-level console probing return the number of
* registered adapters.
*/
if (!(flags & VIO_PROBE_ONLY)) {
for (i = 0; vid_find_adapter(CREATOR_DRIVER_NAME, i) >= 0; i++)
;
return (i);
}
/* Low-level console probing and initialization. */
Fix a typo. Allow for the creator not being stdout.
2003-11-11 07:34:08 +00:00
sc = &creator_softc;
- In creator_configure() when probed for the high-level console return the number of registered adapters instead of determining again whether stdout is a supported card (and which might have failed to attach and register). - Drop creator_set_mode() and move the relevant parts to creator_fill_rect() and creator_putc() respectively. This is a bit cleaner than having to make sure that creator_set_mode() was called before creator_fill_rect() or creator_putc() are used and matches better what Xorg does. - Fix a bug in the handling of the FBIOSCURSOR IOCTL; the code was meant to return ENODEV for all invocations expect when used to disable the cursor and not just when used for enabling the cursor. - In case the adapter is the OFW stdout move its OFW cursor to the start of the last line on halt so OFW output doesn't get intermixed with what FreeBSD left on the screen. With hindsight this is what the faking of a hardware cursor which was removed in the last revision really was about, i.e. to keep the OFW updated about the current cursor position. The new approach however is simpler while producing the same result and doesn't cause the first letter of the OFW output to be turned into a blank and a newline. - Add variable names to the prototypes of creator_cursor_*() which were added in the last revision and list them alphabetically in order to match the style of this file.
2005-06-04 21:15:27 +00:00
if (sc->sc_va.va_flags & V_ADP_REGISTERED)
goto found;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
if ((chosen = OF_finddevice("/chosen")) == -1)
return (0);
if (OF_getprop(chosen, "stdout", &stdout, sizeof(stdout)) == -1)
return (0);
if ((output = OF_instance_to_package(stdout)) == -1)
return (0);
if (OF_getprop(output, "name", buf, sizeof(buf)) == -1)
return (0);
if (strcmp(buf, "SUNW,ffb") == 0 || strcmp(buf, "SUNW,afb") == 0) {
- In creator_configure() when probed for the high-level console return the number of registered adapters instead of determining again whether stdout is a supported card (and which might have failed to attach and register). - Drop creator_set_mode() and move the relevant parts to creator_fill_rect() and creator_putc() respectively. This is a bit cleaner than having to make sure that creator_set_mode() was called before creator_fill_rect() or creator_putc() are used and matches better what Xorg does. - Fix a bug in the handling of the FBIOSCURSOR IOCTL; the code was meant to return ENODEV for all invocations expect when used to disable the cursor and not just when used for enabling the cursor. - In case the adapter is the OFW stdout move its OFW cursor to the start of the last line on halt so OFW output doesn't get intermixed with what FreeBSD left on the screen. With hindsight this is what the faking of a hardware cursor which was removed in the last revision really was about, i.e. to keep the OFW updated about the current cursor position. The new approach however is simpler while producing the same result and doesn't cause the first letter of the OFW output to be turned into a blank and a newline. - Add variable names to the prototypes of creator_cursor_*() which were added in the last revision and list them alphabetically in order to match the style of this file.
2005-06-04 21:15:27 +00:00
sc->sc_flags = CREATOR_CONSOLE;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
if (strcmp(buf, "SUNW,afb") == 0)
sc->sc_flags |= CREATOR_AFB;
sc->sc_node = output;
} else
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
for (i = FFB_DAC; i <= FFB_FBC; i++) {
if (OF_decode_addr(output, i, &space, &addr) != 0)
return (0);
sc->sc_bt[i] = &creator_bst_store[i - FFB_DAC];
sc->sc_bh[i] = sparc64_fake_bustag(space, addr, sc->sc_bt[i]);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
if (creator_init(0, &sc->sc_va, 0) < 0)
return (0);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
found:
/* Return number of found adapters. */
return (1);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
creator_probe(int unit, video_adapter_t **adpp, void *arg, int flags)
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_init(int unit, video_adapter_t *adp, int flags)
{
struct creator_softc *sc;
phandle_t options;
video_info_t *vi;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
char buf[sizeof("screen-#columns")];
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
sc = (struct creator_softc *)adp;
vi = &adp->va_info;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
vid_init_struct(adp, CREATOR_DRIVER_NAME, -1, unit);
Fix whitespace.
2009-04-14 20:51:58 +00:00
if (OF_getprop(sc->sc_node, "height", &sc->sc_height,
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
sizeof(sc->sc_height)) == -1)
return (ENXIO);
if (OF_getprop(sc->sc_node, "width", &sc->sc_width,
sizeof(sc->sc_width)) == -1)
return (ENXIO);
if ((options = OF_finddevice("/options")) == -1)
return (ENXIO);
if (OF_getprop(options, "screen-#rows", buf, sizeof(buf)) == -1)
return (ENXIO);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
vi->vi_height = strtol(buf, NULL, 10);
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
if (OF_getprop(options, "screen-#columns", buf, sizeof(buf)) == -1)
return (ENXIO);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
vi->vi_width = strtol(buf, NULL, 10);
Move the gallant 12 x 22 font data from a .h to a .c so it doesn't need to be compiled into every driver making use of it. Use a const instance of struct gfb_font for this as the font isn't intended to be changed at run-time and in order to accompany the font data with height and width info.
2007-06-16 21:48:50 +00:00
vi->vi_cwidth = gallant12x22.width;
vi->vi_cheight = gallant12x22.height;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
vi->vi_flags = V_INFO_COLOR;
vi->vi_mem_model = V_INFO_MM_OTHER;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
Move the gallant 12 x 22 font data from a .h to a .c so it doesn't need to be compiled into every driver making use of it. Use a const instance of struct gfb_font for this as the font isn't intended to be changed at run-time and in order to accompany the font data with height and width info.
2007-06-16 21:48:50 +00:00
sc->sc_font = gallant12x22.data;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
sc->sc_xmargin = (sc->sc_width - (vi->vi_width * vi->vi_cwidth)) / 2;
sc->sc_ymargin = (sc->sc_height - (vi->vi_height * vi->vi_cheight)) / 2;
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
creator_set_mode(adp, 0);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
if (!(sc->sc_flags & CREATOR_AFB)) {
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE, FFB_DAC_CFG_DID);
if (((FFB_READ(sc, FFB_DAC, FFB_DAC_VALUE) &
FFB_DAC_CFG_DID_PNUM) >> 12) != 0x236e) {
Fix whitespace.
2009-04-14 20:51:58 +00:00
sc->sc_flags |= CREATOR_PAC1;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE, FFB_DAC_CFG_UCTRL);
if (((FFB_READ(sc, FFB_DAC, FFB_DAC_VALUE) &
FFB_DAC_UCTRL_MANREV) >> 8) <= 2)
sc->sc_flags |= CREATOR_CURINV;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
}
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
creator_blank_display(adp, V_DISPLAY_ON);
creator_clear(adp);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
/*
* Setting V_ADP_MODECHANGE serves as hack so creator_set_mode()
* (which will invalidate our caches and restore our settings) is
Fix whitespace.
2009-04-14 20:51:58 +00:00
* called when the X server shuts down. Otherwise screen corruption
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
* happens most of the time.
*/
adp->va_flags |= V_ADP_COLOR | V_ADP_MODECHANGE | V_ADP_BORDER |
V_ADP_INITIALIZED;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
if (vid_register(adp) < 0)
return (ENXIO);
adp->va_flags |= V_ADP_REGISTERED;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_get_info(video_adapter_t *adp, int mode, video_info_t *info)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
bcopy(&adp->va_info, info, sizeof(*info));
return (0);
}
static int
creator_query_mode(video_adapter_t *adp, video_info_t *info)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_set_mode(video_adapter_t *adp, int mode)
{
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
struct creator_softc *sc;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
sc = (struct creator_softc *)adp;
sc->sc_bg_cache = -1;
sc->sc_fg_cache = -1;
sc->sc_fontinc_cache = -1;
sc->sc_fontw_cache = -1;
sc->sc_pmask_cache = -1;
creator_ras_wait(sc);
sc->sc_fifo_cache = 0;
creator_ras_fifo_wait(sc, 2);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_PPC, FBC_PPC_VCE_DIS |
FBC_PPC_TBE_OPAQUE | FBC_PPC_APE_DIS | FBC_PPC_CS_CONST);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_FBC, FFB_FBC_WB_A | FFB_FBC_RB_A |
FFB_FBC_SB_BOTH | FFB_FBC_XE_OFF | FFB_FBC_RGBE_MASK);
return (0);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
Add a font width argument to vi_load_font_t, vi_save_font_t and vi_putm_t and do some preparations for handling 12x22 fonts (currently lots of code implies and/or hardcodes a font width of 8 pixels). This will be required on sparc64 which uses a default font size of 12x22 in order to add font loading and saving support as well as to use a syscons(4)-supplied mouse pointer image. This API breakage is committed now so it can be MFC'ed in time for 6.0 and later on upcoming framebuffer drivers destined for use on sparc64 and which are expected to rely on using font loading internally and on a syscons(4)-supplied mouse pointer image can be easily MFC'ed to RELENG_6 rather than requiring a backport. Tested on: i386, sparc64, make universe MFC after: 1 week
2005-09-28 14:54:07 +00:00
creator_save_font(video_adapter_t *adp, int page, int size, int width,
u_char *data, int c, int count)
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
Add a font width argument to vi_load_font_t, vi_save_font_t and vi_putm_t and do some preparations for handling 12x22 fonts (currently lots of code implies and/or hardcodes a font width of 8 pixels). This will be required on sparc64 which uses a default font size of 12x22 in order to add font loading and saving support as well as to use a syscons(4)-supplied mouse pointer image. This API breakage is committed now so it can be MFC'ed in time for 6.0 and later on upcoming framebuffer drivers destined for use on sparc64 and which are expected to rely on using font loading internally and on a syscons(4)-supplied mouse pointer image can be easily MFC'ed to RELENG_6 rather than requiring a backport. Tested on: i386, sparc64, make universe MFC after: 1 week
2005-09-28 14:54:07 +00:00
creator_load_font(video_adapter_t *adp, int page, int size, int width,
u_char *data, int c, int count)
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_show_font(video_adapter_t *adp, int page)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_save_palette(video_adapter_t *adp, u_char *palette)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_load_palette(video_adapter_t *adp, u_char *palette)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_set_border(video_adapter_t *adp, int border)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
struct creator_softc *sc;
sc = (struct creator_softc *)adp;
creator_fill_rect(adp, border, 0, 0, sc->sc_width, sc->sc_ymargin);
creator_fill_rect(adp, border, 0, sc->sc_height - sc->sc_ymargin,
sc->sc_width, sc->sc_ymargin);
creator_fill_rect(adp, border, 0, 0, sc->sc_xmargin, sc->sc_height);
creator_fill_rect(adp, border, sc->sc_width - sc->sc_xmargin, 0,
sc->sc_xmargin, sc->sc_height);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_save_state(video_adapter_t *adp, void *p, size_t size)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_load_state(video_adapter_t *adp, void *p)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_set_win_org(video_adapter_t *adp, off_t offset)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_read_hw_cursor(video_adapter_t *adp, int *col, int *row)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
*col = 0;
*row = 0;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_set_hw_cursor(video_adapter_t *adp, int col, int row)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_set_hw_cursor_shape(video_adapter_t *adp, int base, int height,
int celsize, int blink)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_blank_display(video_adapter_t *adp, int mode)
{
struct creator_softc *sc;
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
uint32_t v;
int i;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
sc = (struct creator_softc *)adp;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE, FFB_DAC_CFG_TGEN);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
v = FFB_READ(sc, FFB_DAC, FFB_DAC_VALUE);
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
switch (mode) {
case V_DISPLAY_ON:
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
v |= FFB_DAC_CFG_TGEN_VIDE;
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
break;
case V_DISPLAY_BLANK:
case V_DISPLAY_STAND_BY:
case V_DISPLAY_SUSPEND:
v &= ~FFB_DAC_CFG_TGEN_VIDE;
break;
}
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE, FFB_DAC_CFG_TGEN);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
FFB_WRITE(sc, FFB_DAC, FFB_DAC_VALUE, v);
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
for (i = 0; i < 10; i++) {
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE, FFB_DAC_CFG_TGEN);
(void)FFB_READ(sc, FFB_DAC, FFB_DAC_VALUE);
}
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
Update d_mmap() to accept vm_ooffset_t and vm_memattr_t. This replaces d_mmap() with the d_mmap2() implementation and also changes the type of offset to vm_ooffset_t. Purge d_mmap2(). All driver modules will need to be rebuilt since D_VERSION is also bumped. Reviewed by: jhb@ MFC after: Not in this lifetime...
2009-12-29 21:51:28 +00:00
creator_mmap(video_adapter_t *adp, vm_ooffset_t offset, vm_paddr_t *paddr,
int prot, vm_memattr_t *memattr)
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (EINVAL);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_ioctl(video_adapter_t *adp, u_long cmd, caddr_t data)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
struct creator_softc *sc;
struct fbcursor *fbc;
struct fbtype *fb;
sc = (struct creator_softc *)adp;
switch (cmd) {
case FBIOGTYPE:
fb = (struct fbtype *)data;
fb->fb_type = FBTYPE_CREATOR;
fb->fb_height = sc->sc_height;
fb->fb_width = sc->sc_width;
fb->fb_depth = fb->fb_cmsize = fb->fb_size = 0;
break;
case FBIOSCURSOR:
fbc = (struct fbcursor *)data;
- In creator_configure() when probed for the high-level console return the number of registered adapters instead of determining again whether stdout is a supported card (and which might have failed to attach and register). - Drop creator_set_mode() and move the relevant parts to creator_fill_rect() and creator_putc() respectively. This is a bit cleaner than having to make sure that creator_set_mode() was called before creator_fill_rect() or creator_putc() are used and matches better what Xorg does. - Fix a bug in the handling of the FBIOSCURSOR IOCTL; the code was meant to return ENODEV for all invocations expect when used to disable the cursor and not just when used for enabling the cursor. - In case the adapter is the OFW stdout move its OFW cursor to the start of the last line on halt so OFW output doesn't get intermixed with what FreeBSD left on the screen. With hindsight this is what the faking of a hardware cursor which was removed in the last revision really was about, i.e. to keep the OFW updated about the current cursor position. The new approach however is simpler while producing the same result and doesn't cause the first letter of the OFW output to be turned into a blank and a newline. - Add variable names to the prototypes of creator_cursor_*() which were added in the last revision and list them alphabetically in order to match the style of this file.
2005-06-04 21:15:27 +00:00
if (fbc->set & FB_CUR_SETCUR && fbc->enable == 0) {
creator_cursor_enable(sc, 0);
sc->sc_flags &= ~CREATOR_CUREN;
} else
return (ENODEV);
break;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
break;
default:
return (fb_commonioctl(adp, cmd, data));
}
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_clear(video_adapter_t *adp)
{
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
struct creator_softc *sc;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
sc = (struct creator_softc *)adp;
creator_fill_rect(adp, (SC_NORM_ATTR >> 4) & 0xf, 0, 0, sc->sc_width,
sc->sc_height);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_fill_rect(video_adapter_t *adp, int val, int x, int y, int cx, int cy)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
struct creator_softc *sc;
sc = (struct creator_softc *)adp;
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
creator_ras_setpmask(sc, 0xffffffff);
creator_ras_fifo_wait(sc, 2);
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
FFB_WRITE(sc, FFB_FBC, FFB_FBC_ROP, FBC_ROP_NEW);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_DRAWOP, FBC_DRAWOP_RECTANGLE);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
creator_ras_setfg(sc, creator_cmap[val & 0xf]);
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
/*
* Note that at least the Elite3D cards are sensitive to the order
* of operations here.
*/
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
creator_ras_fifo_wait(sc, 4);
FFB_WRITE(sc, FFB_FBC, FFB_FBC_BY, y);
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
FFB_WRITE(sc, FFB_FBC, FFB_FBC_BX, x);
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
FFB_WRITE(sc, FFB_FBC, FFB_FBC_BH, cy);
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
FFB_WRITE(sc, FFB_FBC, FFB_FBC_BW, cx);
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
creator_ras_wait(sc);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_bitblt(video_adapter_t *adp, ...)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_diag(video_adapter_t *adp, int level)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
video_info_t info;
fb_dump_adp_info(adp->va_name, adp, level);
creator_get_info(adp, 0, &info);
fb_dump_mode_info(adp->va_name, adp, &info, level);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_save_cursor_palette(video_adapter_t *adp, u_char *palette)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_load_cursor_palette(video_adapter_t *adp, u_char *palette)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_copy(video_adapter_t *adp, vm_offset_t src, vm_offset_t dst, int n)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_putp(video_adapter_t *adp, vm_offset_t off, u_int32_t p, u_int32_t a,
int size, int bpp, int bit_ltor, int byte_ltor)
{
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
return (ENODEV);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
static int
creator_putc(video_adapter_t *adp, vm_offset_t off, u_int8_t c, u_int8_t a)
{
struct creator_softc *sc;
Move the gallant 12 x 22 font data from a .h to a .c so it doesn't need to be compiled into every driver making use of it. Use a const instance of struct gfb_font for this as the font isn't intended to be changed at run-time and in order to accompany the font data with height and width info.
2007-06-16 21:48:50 +00:00
const uint16_t *p;
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
int row;
int col;
int i;
sc = (struct creator_softc *)adp;
row = (off / adp->va_info.vi_width) * adp->va_info.vi_cheight;
col = (off % adp->va_info.vi_width) * adp->va_info.vi_cwidth;
Move the gallant 12 x 22 font data from a .h to a .c so it doesn't need to be compiled into every driver making use of it. Use a const instance of struct gfb_font for this as the font isn't intended to be changed at run-time and in order to accompany the font data with height and width info.
2007-06-16 21:48:50 +00:00
p = (const uint16_t *)sc->sc_font + (c * adp->va_info.vi_cheight);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
creator_ras_setfg(sc, creator_cmap[a & 0xf]);
creator_ras_setbg(sc, creator_cmap[(a >> 4) & 0xf]);
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
creator_ras_fifo_wait(sc, 1 + adp->va_info.vi_cheight);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
FFB_WRITE(sc, FFB_FBC, FFB_FBC_FONTXY,
((row + sc->sc_ymargin) << 16) | (col + sc->sc_xmargin));
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
creator_ras_setfontw(sc, adp->va_info.vi_cwidth);
creator_ras_setfontinc(sc, 0x10000);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
for (i = 0; i < adp->va_info.vi_cheight; i++) {
FFB_WRITE(sc, FFB_FBC, FFB_FBC_FONT, *p++ << 16);
}
return (0);
}
static int
creator_puts(video_adapter_t *adp, vm_offset_t off, u_int16_t *s, int len)
{
int i;
for (i = 0; i < len; i++) {
Replace explicit calls to video methods with their respective variants implemented with macros. This patch improves code readability. Reasoning behind vidd_* is a sort of "video discipline". List of macros is supposed to be complete--all methods of video_switch ought to have their respective macros from now on. Functionally, this code should be no-op. My intention is to leave current behaviour of touched code as is. No objections: rwatson Silence on: freebsd-current@ Approved by: cognet
2007-12-29 23:26:59 +00:00
vidd_putc(adp, off + i, s[i] & 0xff, (s[i] & 0xff00) >> 8);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
}
Replace explicit calls to video methods with their respective variants implemented with macros. This patch improves code readability. Reasoning behind vidd_* is a sort of "video discipline". List of macros is supposed to be complete--all methods of video_switch ought to have their respective macros from now on. Functionally, this code should be no-op. My intention is to leave current behaviour of touched code as is. No objections: rwatson Silence on: freebsd-current@ Approved by: cognet
2007-12-29 23:26:59 +00:00
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
return (0);
}
static int
creator_putm(video_adapter_t *adp, int x, int y, u_int8_t *pixel_image,
Add a font width argument to vi_load_font_t, vi_save_font_t and vi_putm_t and do some preparations for handling 12x22 fonts (currently lots of code implies and/or hardcodes a font width of 8 pixels). This will be required on sparc64 which uses a default font size of 12x22 in order to add font loading and saving support as well as to use a syscons(4)-supplied mouse pointer image. This API breakage is committed now so it can be MFC'ed in time for 6.0 and later on upcoming framebuffer drivers destined for use on sparc64 and which are expected to rely on using font loading internally and on a syscons(4)-supplied mouse pointer image can be easily MFC'ed to RELENG_6 rather than requiring a backport. Tested on: i386, sparc64, make universe MFC after: 1 week
2005-09-28 14:54:07 +00:00
u_int32_t pixel_mask, int size, int width)
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
{
struct creator_softc *sc;
sc = (struct creator_softc *)adp;
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
if (!(sc->sc_flags & CREATOR_CUREN)) {
creator_cursor_install(sc);
creator_cursor_enable(sc, 1);
sc->sc_flags |= CREATOR_CUREN;
}
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE2, FFB_DAC_CUR_POS);
Add a driver for creator upa frame buffers found in many sparc64 machines. These are fixed resolution and operate only in pixel mode so they present a challenge to syscons (square peg, round hole, etc, etc). The driver provides a video driver interface for syscons and a separate character device for X to mmap. Wherever possible the creator's accelarated graphics functions are used so text mode is very fast. Based roughly on the openbsd driver.
2003-08-24 01:15:40 +00:00
FFB_WRITE(sc, FFB_DAC, FFB_DAC_VALUE2,
((y + sc->sc_ymargin) << 16) | (x + sc->sc_xmargin));
return (0);
}
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
/*
* bus interface
*/
static int
creator_bus_probe(device_t dev)
{
const char *name;
phandle_t node;
int type;
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
name = ofw_bus_get_name(dev);
node = ofw_bus_get_node(dev);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
if (strcmp(name, "SUNW,ffb") == 0) {
if (OF_getprop(node, "board_type", &type, sizeof(type)) == -1)
return (ENXIO);
switch (type & 7) {
case 0x0:
device_set_desc(dev, "Creator");
break;
case 0x3:
device_set_desc(dev, "Creator3D");
break;
default:
return (ENXIO);
}
} else if (strcmp(name, "SUNW,afb") == 0)
device_set_desc(dev, "Elite3D");
else
return (ENXIO);
return (BUS_PROBE_DEFAULT);
}
static int
creator_bus_attach(device_t dev)
{
struct creator_softc *sc;
video_adapter_t *adp;
video_switch_t *sw;
phandle_t node;
int error;
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
int rid;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
int unit;
int i;
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
node = ofw_bus_get_node(dev);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
if ((sc = (struct creator_softc *)vid_get_adapter(vid_find_adapter(
CREATOR_DRIVER_NAME, 0))) != NULL && sc->sc_node == node) {
Fix whitespace.
2009-04-14 20:51:58 +00:00
device_printf(dev, "console\n");
device_set_softc(dev, sc);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
} else {
sc = device_get_softc(dev);
sc->sc_node = node;
}
adp = &sc->sc_va;
/*
* Allocate resources regardless of whether we are the console
* and already obtained the bus tags and handles for the FFB_DAC
* and FFB_FBC register banks in creator_configure() or not so
* the resources are marked as taken in the respective RMAN.
* The supported cards use either 15 (Creator, Elite3D?) or 24
Fix whitespace.
2009-04-14 20:51:58 +00:00
* (Creator3D?) register banks. We make sure that we can also
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
* allocate the resources for at least the FFB_DAC and FFB_FBC
Fix whitespace.
2009-04-14 20:51:58 +00:00
* banks here. We try but don't actually care whether we can
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
* allocate more than these two resources and just limit the
* range accessible via creator_fb_mmap() accordingly.
*/
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
for (i = 0; i < FFB_NREG; i++) {
rid = i;
sc->sc_reg[i] = bus_alloc_resource_any(dev, SYS_RES_MEMORY,
&rid, RF_ACTIVE);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
if (sc->sc_reg[i] == NULL) {
if (i <= FFB_FBC) {
device_printf(dev,
"cannot allocate resources\n");
error = ENXIO;
goto fail;
}
break;
}
sc->sc_bt[i] = rman_get_bustag(sc->sc_reg[i]);
sc->sc_bh[i] = rman_get_bushandle(sc->sc_reg[i]);
}
/*
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
* The XFree86/X.Org sunffb(4) expects to be able to access the
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
* memory spanned by the first and the last resource as one chunk
* via creator_fb_mmap(), using offsets from the first resource,
* even though the backing resources are actually non-continuous.
* So make sure that the memory we provide is at least backed by
* increasing resources.
*/
for (i = 1; i < FFB_NREG && sc->sc_reg[i] != NULL &&
rman_get_start(sc->sc_reg[i]) > rman_get_start(sc->sc_reg[i - 1]);
i++)
;
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
sc->sc_reg_size = rman_get_end(sc->sc_reg[i - 1]) -
rman_get_start(sc->sc_reg[0]) + 1;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
if (!(sc->sc_flags & CREATOR_CONSOLE)) {
if ((sw = vid_get_switch(CREATOR_DRIVER_NAME)) == NULL) {
device_printf(dev, "cannot get video switch\n");
error = ENODEV;
goto fail;
}
/*
* During device configuration we don't necessarily probe
* the adapter which is the console first so we can't use
Fix whitespace.
2009-04-14 20:51:58 +00:00
* the device unit number for the video adapter unit. The
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
* worst case would be that we use the video adapter unit
Fix whitespace.
2009-04-14 20:51:58 +00:00
* 0 twice. As it doesn't really matter which unit number
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
* the corresponding video adapter has just use the next
* unused one.
*/
for (i = 0; i < devclass_get_maxunit(creator_devclass); i++)
if (vid_find_adapter(CREATOR_DRIVER_NAME, i) < 0)
break;
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
if (strcmp(ofw_bus_get_name(dev), "SUNW,afb") == 0)
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
sc->sc_flags |= CREATOR_AFB;
if ((error = sw->init(i, adp, 0)) != 0) {
device_printf(dev, "cannot initialize adapter\n");
Fix whitespace.
2009-04-14 20:51:58 +00:00
goto fail;
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
}
}
if (bootverbose) {
if (sc->sc_flags & CREATOR_PAC1)
device_printf(dev,
"BT9068/PAC1 RAMDAC (%s cursor control)\n",
sc->sc_flags & CREATOR_CURINV ? "inverted" :
"normal");
else
device_printf(dev, "BT498/PAC2 RAMDAC\n");
}
device_printf(dev, "resolution %dx%d\n", sc->sc_width, sc->sc_height);
unit = device_get_unit(dev);
sc->sc_si = make_dev(&creator_fb_devsw, unit, UID_ROOT, GID_WHEEL,
0600, "fb%d", unit);
sc->sc_si->si_drv1 = sc;
EVENTHANDLER_REGISTER(shutdown_final, creator_shutdown, sc,
SHUTDOWN_PRI_DEFAULT);
return (0);
fail:
for (i = 0; i < FFB_NREG && sc->sc_reg[i] != NULL; i++)
Rototill the sparc64 nexus(4) (actually this brings in the code the sun4v nexus(4) in turn is based on): o Change nexus(4) to manage the resources of its children so the respective device drivers don't need to figure them out of OFW themselves. o Change nexus(4) to provide the ofw_bus KOBJ interface instead of using IVARs for supplying the OFW node and the subset of standard properties of its children. Together with the previous change this also allows to fully take advantage of newbus in that drivers like fhc(4), which attach on multiple parent busses, no longer require different bus front-ends as obtaining the OFW node and properties as well as resource allocation works the same for all supported busses. As such this change also is part 4/4 of allowing creator(4) to work in USIII-based machines as it allows this driver to attach on both nexus(4) and upa(4). On the other hand removing these IVARs breaks API compatibility with the powerpc nexus(4) but which isn't that bad as a) sparc64 currently doesn't share any device driver hanging off of nexus(4) with powerpc and b) they were no longer compatible regarding OFW-related extensions at the pci(4) level since quite some time. o Provide bus_get_dma_tag methods in nexus(4) and its children in order to handle DMA tags in a hierarchical way and get rid of the sparc64_root_dma_tag kludge. Together with the previous two items this changes also allows to completely get rid of the nexus(4) IVAR interface. It also includes: - pushing the constraints previously specified by the nexus_dmatag down into the DMA tags of psycho(4) and sbus(4) as it's their IOMMUs which induce these restrictions (and nothing at the nexus(4) or anything that would warrant specifying them there), - fixing some obviously wrong constraints of the psycho(4) and sbus(4) DMA tags, which happened to not actually be used with the sparc64_root_dma_tag kludge in place and therefore didn't cause problems so far, - replacing magic constants for constraints with macros as far as it is obvious as to where they come from. This doesn't include taking advantage of the newbus way to get the parent DMA tags implemented by this change in order to divorce the IOTSBs of the PCI and SBus IOMMUs or for implementing the workaround for the DMA sync bug in Sabre (and Tomatillo) bridges, yet, though. o Get rid of the notion that nexus(4) (mostly) reflects an UPA bus by replacing ofw_upa.h and with ofw_nexus.h (which was repo-copied from ofw_upa.h) and renaming its content, which actually applies to all of Fireplane/Safari, JBus and UPA (in the host bus case), as appropriate. o Just use M_DEVBUF instead of a separate M_NEXUS malloc type for allocating the device info for the children of nexus(4). This is done in order to not need to export M_NEXUS when deriving drivers for subordinate busses from the nexus(4) class. o Use the DEFINE_CLASS_0() macro to declare the nexus(4) driver so we can derive subclasses from it. o Const'ify the nexus_excl_name and nexus_excl_type arrays as well as add 'associations' and 'rsc', which are pseudo-devices without resources and therefore of no real interest for nexus(4), to the former. o Let the nexus(4) device memory rman manage the entire 64-bit address space instead of just the UPA_MEMSTART to UPA_MEMEND subregion as Fireplane/Safari- and JBus-based machines use multiple ranges, which can't be as easily divided as in the case of UPA (limiting the address space only served for sanity checking anyway). o Use M_WAITOK instead of M_NOWAIT when allocating the device info for children of nexus(4) in order to give one less opportunity for adding devices to nexus(4) to fail. o While adapting the drivers affected by the above nexus(4) changes, change them to take advantage of rman_get_rid() instead of caching the RIDs assigned to allocated resources, now that the RIDs of resources are correctly set. o In iommu(4) and nexus(4) replace hard-coded functions names, which actually became outdated in several places, in panic strings and status massages with __func__. [1] o Use driver_filter_t in prototypes where appropriate. o Add my copyright to creator(4), fhc(4), nexus(4), psycho(4) and sbus(4) as I changed considerable amounts of these drivers as well as added a bunch of new features, workarounds for silicon bugs etc. o Fix some white space nits. Due to lack of access to Exx00 hardware, these changes, i.e. central(4) and fhc(4), couldn't be runtime tested on such a machine. Exx00 are currently reported to panic before trying to attach nexus(4) anyway though. PR: 76052 [1] Approved by: re (kensmith)
2007-03-07 21:13:51 +00:00
bus_release_resource(dev, SYS_RES_MEMORY,
rman_get_rid(sc->sc_reg[i]), sc->sc_reg[i]);
Fix whitespace.
2009-04-14 20:51:58 +00:00
return (error);
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
}
/*
* /dev/fb interface
*/
static int
creator_fb_open(struct cdev *dev, int flags, int mode, struct thread *td)
{
return (0);
}
static int
creator_fb_close(struct cdev *dev, int flags, int mode, struct thread *td)
{
return (0);
}
static int
creator_fb_ioctl(struct cdev *dev, u_long cmd, caddr_t data, int flags,
struct thread *td)
{
struct creator_softc *sc;
sc = dev->si_drv1;
return (creator_ioctl(&sc->sc_va, cmd, data));
}
static int
Update d_mmap() to accept vm_ooffset_t and vm_memattr_t. This replaces d_mmap() with the d_mmap2() implementation and also changes the type of offset to vm_ooffset_t. Purge d_mmap2(). All driver modules will need to be rebuilt since D_VERSION is also bumped. Reviewed by: jhb@ MFC after: Not in this lifetime...
2009-12-29 21:51:28 +00:00
creator_fb_mmap(struct cdev *dev, vm_ooffset_t offset, vm_paddr_t *paddr,
int prot, vm_memattr_t *memattr)
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
{
struct creator_softc *sc;
int i;
/*
* NB: This is a special implementation based on the /dev/fb
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
* requirements of the XFree86/X.Org sunffb(4).
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
*/
sc = dev->si_drv1;
for (i = 0; i < CREATOR_FB_MAP_SIZE; i++) {
if (offset >= creator_fb_map[i].virt &&
offset < creator_fb_map[i].virt + creator_fb_map[i].size) {
Fix whitespace.
2009-04-14 20:51:58 +00:00
offset += creator_fb_map[i].phys -
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
creator_fb_map[i].virt;
- Change some softc members to be unsigned where more appropriate. - Add some missing const. - Move the size of the window spun by the registers to the softc as neither using va_mem_size for this nor va_mem_base for the start of the bus addresses is appropriate. MFC after: 1 week
2009-04-28 20:49:47 +00:00
if (offset >= sc->sc_reg_size)
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
return (EINVAL);
*paddr = sc->sc_bh[0] + offset;
return (0);
}
}
return (EINVAL);
}
/*
* internal functions
*/
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
static void
creator_cursor_enable(struct creator_softc *sc, int onoff)
{
int v;
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE2, FFB_DAC_CUR_CTRL);
if (sc->sc_flags & CREATOR_CURINV)
v = onoff ? FFB_DAC_CUR_CTRL_P0 | FFB_DAC_CUR_CTRL_P1 : 0;
else
v = onoff ? 0 : FFB_DAC_CUR_CTRL_P0 | FFB_DAC_CUR_CTRL_P1;
FFB_WRITE(sc, FFB_DAC, FFB_DAC_VALUE2, v);
}
static void
creator_cursor_install(struct creator_softc *sc)
{
int i, j;
creator_cursor_enable(sc, 0);
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE2, FFB_DAC_CUR_COLOR1);
FFB_WRITE(sc, FFB_DAC, FFB_DAC_VALUE2, 0xffffff);
FFB_WRITE(sc, FFB_DAC, FFB_DAC_VALUE2, 0x0);
for (i = 0; i < 2; i++) {
FFB_WRITE(sc, FFB_DAC, FFB_DAC_TYPE2,
i ? FFB_DAC_CUR_BITMAP_P0 : FFB_DAC_CUR_BITMAP_P1);
for (j = 0; j < 64; j++) {
FFB_WRITE(sc, FFB_DAC, FFB_DAC_VALUE2,
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
*(const uint32_t *)(&creator_mouse_pointer[j][0]));
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
FFB_WRITE(sc, FFB_DAC, FFB_DAC_VALUE2,
- Declare lookup tables etc. const. - Let creator_bitblt() return ENODEV as it's not implemented (missed in sys/dev/fb/creator.c rev. 1.6). - As a speed optimization inline the creator_ras_wait() etc. helper functions and also cache setting the font increment, font width and plane mask. [1] - I got the meaning of V_DISPLAY_BLANK wrong, it's blank like turn off and not blank like turn on and clear the screen. So move clearing the screen to creator_clear() were it hopefully belongs. - Properly implement V_DISPLAY_BLANK, V_DISPLAY_STAND_BY and V_DISPLAY_SUSPEND. This makes blank_saver.ko and green_saver.ko work. [1] - Change the order of operations in creator_fill_rect(), i.e. write y before x and cy before cx. This fixes drawing the top part of the border with Elite3D cards when switching from Xorg to a VTY. - Move setting the chip configuration we use and invalidating the cache variables to creator_set_mode() and set the V_ADP_MODECHANGE flag. This causes creator_set_mode() to be called when the X server shuts down which fixes the screen corruption caused most of the time by Xorg not restoring the original configuration present at startup. Inspired by/based on: Xorg [1] Approved by: re (scottl)
2005-07-10 11:16:34 +00:00
*(const uint32_t *)(&creator_mouse_pointer[j][4]));
o creator(4): - Use register macros instead of magic values in the code. [1] - Check the return values of OF_getprop() and other stuff that actually can fail. - Let the unimplemented video driver methods return ENODEV rather than 0 so other code isn't tricked into thinking a certain operation was successfull. In case of e.g. the video driver creator_ioctl() this caused vidcontrol(1) to return random garbage information. Remove the TODO macros in the unimplemented video driver methods which did a printf("%s: unimplemented\n", __func__). Under certain circumstances these managed to invoke a printf() when a low-level console device wasn't attached, yet, causing a Fast Data Access MMU Miss. These macros were only really usefull for development anyway. - Set the struct video_adapter and struct video_info va_flags and vi_flags etc. as appropriate. - In creator_configure() don't rely on hitting the node which is the chosen console device first when searching the OFW tree for adapters compatible with this driver. Instead just check whether the chosen console device is a viable target for this driver. Targets that are not the console (including additional cards in multi-head configs) will be attached through creator_upa_attach(). I think this how the code in creator_configure() was actually meant to work. Honour the VIO_PROBE_ONLY flag and don't initialise and register the console device twice when creator_configure() is called a second time during sc_probe_unit(). Let creator_configure() return the number of the found adapters, i.e. 1 in case probing succeeds, as it's expected. The return values of video adapter configure functions however currently aren't checked so this doesn't make a difference at the moment. - In creator_upa_attach() don't rely on probing and attaching the adapter which is the console first, in case there are multiple adpaters and one of them is the console this could lead into using the video adapter unit 0 twice. - Make the check for DACs with inverted cursor control a bit more precise and actually honour that information when turning the cursor on or off. Add a helper function creator_cursor_enable() for this in order to keep code duplication low. [1] - Don't bother with faking a hardware cursor in case a device is the console. Apparently this was meant to start kernel output right after where the firmware left. In general this isn't worth the fuzz and also had no real effect as creator_set_mode() did clear the screen in any case, not just in case a device was not the console. - Implement creator_fill_rect() and use it to actually blank the display in creator_blank_display() when the mode is V_DISPLAY_BLANK, moving blanking the display out of creator_set_mode(). Use it also to implement creator_set_border() so the border can be re-drawn when switching to a VTY from X, exiting X, etc. (which leaves us with a black border most of the time). - Implement the video driver creator_ioctl(), moving the implementation of the IOCTL interface from the fbN CDEV version of creator_ioctl() into the video driver version and use the latter to implement the former. Use fb_commonioctl() to handle most of the FBIO IOCTLs. This gives programs like vidcontrol(1) which use the video driver creator_ioctl() a chance of working. Implement turning off the cursor via the FBIOSCURSOR IOCTL, which Xorg uses to in order to inform the OS that it's taking over the cursor. In creator_putm() check whether the cursor is enabled and (re-)install it if necessary, moving installing the cursor out of creator_init() and into a helper function creator_cursor_install(). This fixes the missing mouse pointer when switching to a VTY from X, exiting X, etc. - Some clean-up (remove unused/useless code, etc.). o sparc64/creator/creator_upa.c / sparc64/sparc64/sc_machdep.c: - Attach syscons(4) as an own pseudo-device on the nexus rather than directly in creator_upa_attach(), similiar to attaching syscons(4) as a pseudo-device on isa(4) on other archs. This makes it a whole lot easier to do the right thing in multi-head configs, especially with different types of graphics adapters. [2] - Set SC_AUTODETECT_KBD by default so USB keyboards work out of the box. [2] Based on/obtained from: Xorg 'ffb' driver [1] Based on/obtained from: FreeBSD/powerpc [2]
2005-05-21 20:38:26 +00:00
}
}
}
- Merge sys/sparc64/creator/creator_upa.c into sys/dev/fb/creator.c. The separate bus front-end was inherited from the OpenBSD creator(4), which at that time had a mainbus(4) (for USI/II machines, which use an UPA interconnection bus as the nexus) and an upa(4) (for USIII machines, which use a subordinate/slave UPA bus hanging off from the Fireplane/Safari interconnection bus) front-end. With FreeBSD and newbus there is/will be no need to have two separate bus front-ends for these busses, so we can easily coallapse the shared front-end and the back-end into a single source file (note that the FreeBSD creator_upa.c was misnomer anyway; based on what it actually attached to that should have been creator_nexus.c), actually OpenBSD meanwhile also has moved to a shared front-end and a single source file. Due to the low-level console support creator.c also wasn't free from bus related things before. While at it, also split sys/sparc64/creator/creator.h into a sys/dev/fb/creatorreg.h that only contains register macros and move the structures to the top of sys/dev/fb/creator.c as suggested by style(9) so creator(4) is no longer scattered over two directories. - Use OF_decode_addr()/sparc64_fake_bustag() to obtain the bus tags and handles for the low-level console support instead of hardcoding support for AFB/FFB hanging off from nexus(4) only. This is part 2/4 of allowing creator(4) to work in USIII machines (which have a UPA bus hanging off from the Fireplane/Safari bus reflected by the nexus), which already makes it work as the low-level console there. - Allocate resources in the bus attach routine regardless of whether creator(4) is used as for the low-level console and thus the required bus tags and handles have been already obtained or not so the resources are marked as taken in the respective RMAN. - For both obtaining the bus tags and handles for the low-level console support as well as allocating the corresponding resources in the regular bus attach routine don't bother to get all for the maximum of 24 register banks but only (for) the two tag/handle pairs required for providing the video interface for syscons(4) support. If we can't allocate the rest of them just limit the memory range accessible via creator_fb_mmap() accordingly. - Sanity check the memory range spanned by the first and last resources and the resources in between as far as possible, as the XFree86/Xorg sunffb(4) expects to be able to access the whole region, even though the backing resources are actually non-continuous. Limit and check the memory range accessible via creator_fb_mmap() accordingly. - Reduce the size of buffers for OFW properties to what they actually need to hold. - Rename some tables to creator_<foo> for consistency. - Also for the sizes in the creator_fb_mmap() mapping table entries use macros for consistency, add macros for the remaining register banks for completeness.
2007-01-16 21:08:22 +00:00
static void
creator_shutdown(void *xsc)
{
struct creator_softc *sc = xsc;
creator_cursor_enable(sc, 0);
/*
* In case this is the console set the cursor of the stdout
* instance to the start of the last line so OFW output ends
* up beneath what FreeBSD left on the screen.
*/
if (sc->sc_flags & CREATOR_CONSOLE) {
OF_interpret("stdout @ is my-self 0 to column#", 0);
OF_interpret("stdout @ is my-self #lines 1 - to line#", 0);
}
}
Reference in New Issue Copy Permalink