d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
b0643f118c
freebsd-nq/sys/compat/ndis/subr_ndis.c

2418 lines
57 KiB
C
Raw Normal View History

Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
/*
* Copyright (c) 2003
* Bill Paul <wpaul@windriver.com>. 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Bill Paul.
* 4. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY Bill Paul 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 Bill Paul OR THE VOICES IN HIS HEAD
* 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
/*
* This file implements a translation layer between the BSD networking
* infrasturcture and Windows(R) NDIS network driver modules. A Windows
* NDIS driver calls into several functions in the NDIS.SYS Windows
* kernel module and exports a table of functions designed to be called
* by the NDIS subsystem. Using the PE loader, we can patch our own
* versions of the NDIS routines into a given Windows driver module and
* convince the driver that it is in fact running on Windows.
*
* We provide a table of all our implemented NDIS routines which is patched
* into the driver object code. All our exported routines must use the
* _stdcall calling convention, since that's what the Windows object code
* expects.
*/
#include <sys/param.h>
#include <sys/types.h>
#include <sys/errno.h>
#include <sys/callout.h>
#include <sys/kernel.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/socket.h>
#include <sys/sysctl.h>
Implement some more NDIS and ntoskrnl API calls: subr_ndis.c: NdisGetCurrentSystemTime() which, according to the Microsoft documentation returns "the number of 100 nanosecond intervals since January 1, 1601." I have no idea what's so special about that epoch or why they chose 100 nanosecond ticks. I don't know the proper offset to convert nanotime() from the UNIX epoch to January 1, 1601, so for now I'm just doing the unit convertion to 100s of nanoseconds. subr_ntoskrnl.c: memcpy(), memset(), ExInterlockedPopEntrySList(), ExInterlockedPushEntrySList(). The latter two are different from InterlockedPopEntrySList() and InterlockedPushEntrySList() in that they accept a spinlock to hold while executing, whereas the non-Ex routines use a lock internal to ntoskrnl. I also modified ExInitializePagedLookasideList() and ExInitializeNPagedLookasideList() to initialize mutex locks within the lookaside structures. It seems that in NDIS 5.0, the lookaside allocate/free routines ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(), which require the use of the per-lookaside spinlock, whereas in NDIS 5.1, the per-lookaside spinlock is deprecated. We need to support both cases. Note that I appear to be doing something wrong with ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(): they don't appear to obtain proper pointers to their arguments, so I'm probably doing something wrong in terms of their calling convention (they're declared to be FASTCALL in Widnows, and I'm not sure what that means for gcc). It happens that in my stub lookaside implementation, they don't need to do any work anyway, so for now I've hacked them to always return NULL, which avoids corrupting the stack. I need to do this right though.
2003-12-12 22:35:13 +00:00
#include <sys/timespec.h>
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
#include <sys/smp.h>
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
#include <net/if.h>
#include <net/if_arp.h>
#include <net/ethernet.h>
#include <net/if_dl.h>
#include <net/if_media.h>
#include <machine/bus_memio.h>
#include <machine/bus_pio.h>
#include <machine/bus.h>
#include <machine/resource.h>
#include <sys/bus.h>
#include <sys/rman.h>
#include <machine/stdarg.h>
Big round of updates: - Make ndis_get_info()/ndis_set_info() sleep on the setdone/getdone routines if they get back NDIS_STATUS_PENDING. - Add a bunch of net80211 support so that 802.11 cards can be twiddled with ifconfig. This still needs more work and is not guaranteed to work for everyone. It works on my 802.11b/g card anyway. The problem here is Microsoft doesn't provide a good way to a) learn all the rates that a card supports (if it has more than 8, you're kinda hosed) and b) doesn't provide a good way to distinguish between 802.11b, 802.11b/g an 802.11a/b/g cards, so you sort of have to guess. Setting the SSID and switching between infrastructure/adhoc modes should work. WEP still needs to be implemented. I can't find any API for getting/setting the channel other than the registry/sysctl keys.
2003-12-21 00:00:08 +00:00
#include <net80211/ieee80211_var.h>
#include <net80211/ieee80211_ioctl.h>
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
#include <dev/pci/pcireg.h>
#include <dev/pci/pcivar.h>
#include <compat/ndis/pe_var.h>
#include <compat/ndis/resource_var.h>
Correct the implementation of NDIS_BUFFER_TO_SPAN_PAGES().
2003-12-13 09:07:35 +00:00
#include <compat/ndis/ntoskrnl_var.h>
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
#include <compat/ndis/ndis_var.h>
#include <compat/ndis/cfg_var.h>
#include <dev/if_ndis/if_ndisvar.h>
#define __stdcall __attribute__((__stdcall__))
#define FUNC void(*)(void)
static struct mtx ndis_interlock;
static int ndis_inits = 0;
__stdcall static void ndis_initwrap(ndis_handle,
ndis_driver_object *, void *, void *);
__stdcall static ndis_status ndis_register_miniport(ndis_handle,
ndis_miniport_characteristics *, int);
__stdcall static ndis_status ndis_malloc_withtag(void **, uint32_t, uint32_t);
__stdcall static ndis_status ndis_malloc(void **,
uint32_t, uint32_t, ndis_physaddr);
__stdcall static void ndis_free(void *, uint32_t, uint32_t);
__stdcall static ndis_status ndis_setattr_ex(ndis_handle, ndis_handle,
uint32_t, uint32_t, ndis_interface_type);
__stdcall static void ndis_open_cfg(ndis_status *, ndis_handle *, ndis_handle);
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void ndis_open_cfgbyidx(ndis_status *, ndis_handle,
uint32_t, ndis_unicode_string *, ndis_handle *);
__stdcall static void ndis_open_cfgbyname(ndis_status *, ndis_handle,
ndis_unicode_string *, ndis_handle *);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
static ndis_status ndis_encode_parm(ndis_miniport_block *,
struct sysctl_oid *, ndis_parm_type, ndis_config_parm **);
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
static ndis_status ndis_decode_parm(ndis_miniport_block *,
ndis_config_parm *, char *);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void ndis_read_cfg(ndis_status *, ndis_config_parm **,
ndis_handle, ndis_unicode_string *, ndis_parm_type);
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void ndis_write_cfg(ndis_status *, ndis_handle,
ndis_unicode_string *, ndis_config_parm *);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void ndis_close_cfg(ndis_handle);
__stdcall static void ndis_create_lock(ndis_spin_lock *);
__stdcall static void ndis_destroy_lock(ndis_spin_lock *);
__stdcall static void ndis_lock(ndis_spin_lock *);
__stdcall static void ndis_unlock(ndis_spin_lock *);
__stdcall static uint32_t ndis_read_pci(ndis_handle, uint32_t,
uint32_t, void *, uint32_t);
__stdcall static uint32_t ndis_write_pci(ndis_handle, uint32_t,
uint32_t, void *, uint32_t);
static void ndis_syslog(ndis_handle, ndis_error_code, uint32_t, ...);
static void ndis_map_cb(void *, bus_dma_segment_t *, int, int);
__stdcall static void ndis_vtophys_load(ndis_handle, ndis_buffer *,
uint32_t, uint8_t, ndis_paddr_unit *, uint32_t *);
__stdcall static void ndis_vtophys_unload(ndis_handle, ndis_buffer *, uint32_t);
__stdcall static void ndis_create_timer(ndis_miniport_timer *, ndis_handle *,
ndis_timer_function, void *);
static void ndis_timercall(void *);
__stdcall static void ndis_set_timer(ndis_miniport_timer *, uint32_t);
static void ndis_tick(void *);
__stdcall static void ndis_set_periodic_timer(ndis_miniport_timer *, uint32_t);
__stdcall static void ndis_cancel_timer(ndis_miniport_timer *, uint8_t *);
__stdcall static void ndis_query_resources(ndis_status *, ndis_handle,
ndis_resource_list *, uint32_t *);
__stdcall static ndis_status ndis_register_ioport(void **,
ndis_handle, uint32_t, uint32_t);
__stdcall static void ndis_deregister_ioport(ndis_handle,
uint32_t, uint32_t, void *);
__stdcall static void ndis_read_netaddr(ndis_status *, void **,
uint32_t *, ndis_handle);
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
__stdcall static ndis_status ndis_mapreg_cnt(uint32_t, uint32_t *);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static ndis_status ndis_alloc_mapreg(ndis_handle,
uint32_t, uint8_t, uint32_t, uint32_t);
__stdcall static void ndis_free_mapreg(ndis_handle);
static void ndis_mapshared_cb(void *, bus_dma_segment_t *, int, int);
__stdcall static void ndis_alloc_sharedmem(ndis_handle, uint32_t,
uint8_t, void **, ndis_physaddr *);
__stdcall static void ndis_alloc_sharedmem_async(ndis_handle,
uint32_t, uint8_t, void *);
__stdcall static void ndis_free_sharedmem(ndis_handle, uint32_t,
uint8_t, void *, ndis_physaddr);
__stdcall static ndis_status ndis_map_iospace(void **, ndis_handle,
ndis_physaddr, uint32_t);
__stdcall static void ndis_unmap_iospace(ndis_handle, void *, uint32_t);
__stdcall static uint32_t ndis_cachefill(void);
__stdcall static uint32_t ndis_dma_align(ndis_handle);
__stdcall static ndis_status ndis_init_sc_dma(ndis_handle,
uint8_t, uint32_t);
__stdcall static void ndis_alloc_packetpool(ndis_status *,
ndis_handle *, uint32_t, uint32_t);
__stdcall static void ndis_ex_alloc_packetpool(ndis_status *,
ndis_handle *, uint32_t, uint32_t, uint32_t);
__stdcall static uint32_t ndis_packetpool_use(ndis_handle);
__stdcall static void ndis_free_packetpool(ndis_handle);
__stdcall static void ndis_alloc_packet(ndis_status *,
ndis_packet **, ndis_handle);
__stdcall static void ndis_release_packet(ndis_packet *);
__stdcall static void ndis_unchain_headbuf(ndis_packet *, ndis_buffer **);
Now that I finally have power back, implement a couple more NDIS API routines: NdisUnchainBufferAtBack(), NdisGetFirstBufferFromPacketSafe() and NdisGetFirstBufferFromPacket(). This should bring us a little closer to getting the Intel centrino wireless NIC to work. Note: I have not actually tested these additions since I don't have a driver that calls them, however they're pretty simple, and one of them is taken pretty much directly from the Windows ndis.h header file, so I'm fairly confident they work, but disclaimers apply.
2003-12-22 08:24:32 +00:00
__stdcall static void ndis_unchain_tailbuf(ndis_packet *, ndis_buffer **);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void ndis_alloc_bufpool(ndis_status *,
ndis_handle *, uint32_t);
__stdcall static void ndis_free_bufpool(ndis_handle);
__stdcall static void ndis_alloc_buf(ndis_status *, ndis_buffer **,
ndis_handle, void *, uint32_t);
__stdcall static void ndis_release_buf(ndis_buffer *);
__stdcall static void ndis_query_buf(ndis_buffer *, void **, uint32_t *);
__stdcall static void ndis_query_buf_safe(ndis_buffer *, void **,
uint32_t *, uint32_t);
__stdcall static void ndis_adjust_buflen(ndis_buffer *, int);
__stdcall static uint32_t ndis_interlock_inc(uint32_t *);
__stdcall static uint32_t ndis_interlock_dec(uint32_t *);
__stdcall static void ndis_init_event(ndis_event *);
__stdcall static void ndis_set_event(ndis_event *);
__stdcall static void ndis_reset_event(ndis_event *);
__stdcall static uint8_t ndis_wait_event(ndis_event *, uint32_t);
__stdcall static ndis_status ndis_unicode2ansi(ndis_ansi_string *,
ndis_unicode_string *);
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static ndis_status ndis_ansi2unicode(ndis_unicode_string *,
ndis_ansi_string *);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static ndis_status ndis_assign_pcirsrc(ndis_handle,
uint32_t, ndis_resource_list **);
__stdcall static ndis_status ndis_register_intr(ndis_miniport_interrupt *,
ndis_handle, uint32_t, uint32_t, uint8_t,
uint8_t, ndis_interrupt_mode);
__stdcall static void ndis_deregister_intr(ndis_miniport_interrupt *);
__stdcall static void ndis_register_shutdown(ndis_handle, void *,
ndis_shutdown_handler);
__stdcall static void ndis_deregister_shutdown(ndis_handle);
__stdcall static uint32_t ndis_numpages(ndis_buffer *);
Implement NdisGetBufferPhysicalArraySize(), which apparently is a synonym for NDIS_BUFFER_TO_SPAN_PAGES().
2003-12-16 18:56:33 +00:00
__stdcall static void ndis_buf_physpages(ndis_buffer *, uint32_t *);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void ndis_query_bufoffset(ndis_buffer *,
uint32_t *, uint32_t *);
__stdcall static void ndis_sleep(uint32_t);
__stdcall static uint32_t ndis_read_pccard_amem(ndis_handle,
uint32_t, void *, uint32_t);
__stdcall static uint32_t ndis_write_pccard_amem(ndis_handle,
uint32_t, void *, uint32_t);
__stdcall static ndis_list_entry *ndis_insert_head(ndis_list_entry *,
ndis_list_entry *, ndis_spin_lock *);
__stdcall static ndis_list_entry *ndis_remove_head(ndis_list_entry *,
ndis_spin_lock *);
__stdcall static ndis_list_entry *ndis_insert_tail(ndis_list_entry *,
ndis_list_entry *, ndis_spin_lock *);
__stdcall static uint8_t ndis_sync_with_intr(ndis_miniport_interrupt *,
void *, void *);
Implement some more NDIS and ntoskrnl API calls: subr_ndis.c: NdisGetCurrentSystemTime() which, according to the Microsoft documentation returns "the number of 100 nanosecond intervals since January 1, 1601." I have no idea what's so special about that epoch or why they chose 100 nanosecond ticks. I don't know the proper offset to convert nanotime() from the UNIX epoch to January 1, 1601, so for now I'm just doing the unit convertion to 100s of nanoseconds. subr_ntoskrnl.c: memcpy(), memset(), ExInterlockedPopEntrySList(), ExInterlockedPushEntrySList(). The latter two are different from InterlockedPopEntrySList() and InterlockedPushEntrySList() in that they accept a spinlock to hold while executing, whereas the non-Ex routines use a lock internal to ntoskrnl. I also modified ExInitializePagedLookasideList() and ExInitializeNPagedLookasideList() to initialize mutex locks within the lookaside structures. It seems that in NDIS 5.0, the lookaside allocate/free routines ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(), which require the use of the per-lookaside spinlock, whereas in NDIS 5.1, the per-lookaside spinlock is deprecated. We need to support both cases. Note that I appear to be doing something wrong with ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(): they don't appear to obtain proper pointers to their arguments, so I'm probably doing something wrong in terms of their calling convention (they're declared to be FASTCALL in Widnows, and I'm not sure what that means for gcc). It happens that in my stub lookaside implementation, they don't need to do any work anyway, so for now I've hacked them to always return NULL, which avoids corrupting the stack. I need to do this right though.
2003-12-12 22:35:13 +00:00
__stdcall static void ndis_time(uint64_t *);
- Add stubs for Ndis*File() functions - Fix ndis_time(). - Implement NdisGetSystemUpTime(). - Implement RtlCopyUnicodeString() and RtlUnicodeStringToAnsiString(). - In ndis_getstate_80211(), use sc->ndis_link to determine connect status. Submitted by: Brian Feldman <green@freebsd.org>
2003-12-25 00:40:02 +00:00
__stdcall static void ndis_uptime(uint32_t *);
subr_ndis.c: - fix ndis_time() so that it returns a time based on the proper epoch (wacky though it may be) - implement NdisInitializeString() and NdisFreeString(), and add stub for NdisMRemoveMiniport() ntoskrnl_var.h: - add missing member to the general_lookaside struct (gl_listentry) subr_ntoskrnl.c: - Fix arguments to the interlocked push/pop routines: 'head' is an slist_header *, not an slist_entry * - Kludge up _fastcall support for the push/pop routines. The _fastcall convention is similar to _stdcall, except the first two available DWORD-sized arguments are passed in %ecx and %edx, respectively. One kludge for this __attribute__ ((regparm(3))), however this isn't entirely right, as it assumes %eax, %ecx and %edx will be used (regparm(2) assumes %eax and %edx). Another kludge is to declare the two fastcall-ed args as local register variables and explicitly assign them to %ecx and %edx, but experimentation showed that gcc would not guard %ecx and %edx against being clobbered. Thus, I came up with a 3rd kludge, which is to use some inline assembly of the form: void *arg1; void *arg2; __asm__("movl %%ecx, %%ecx" : "=c" (arg1)); __asm__("movl %%edx, %%edx" : "=d" (arg2)); This lets gcc know that we're going to reference %ecx and %edx and that it should make an effort not to let it get trampled. This wastes an instruction (movl %reg, %reg is a no-op) but insures proper behavior. It's possible there's a better way to do this though: this is the first time I've used inline assembler in this fashion. The above fixes to ntoskrnl_var.h an subr_ntoskrnl.c make lookaside lists work for the two drivers I have that use them, one of which is an NDIS 5.0 miniport and another which is 5.1.
2003-12-13 07:41:12 +00:00
__stdcall static void ndis_init_string(ndis_unicode_string **, char *);
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void ndis_init_ansi_string(ndis_ansi_string *, char *);
Implement NdisInitUnicodeString().
2003-12-28 21:36:03 +00:00
__stdcall static void ndis_init_unicode_string(ndis_unicode_string *,
uint16_t *);
subr_ndis.c: - fix ndis_time() so that it returns a time based on the proper epoch (wacky though it may be) - implement NdisInitializeString() and NdisFreeString(), and add stub for NdisMRemoveMiniport() ntoskrnl_var.h: - add missing member to the general_lookaside struct (gl_listentry) subr_ntoskrnl.c: - Fix arguments to the interlocked push/pop routines: 'head' is an slist_header *, not an slist_entry * - Kludge up _fastcall support for the push/pop routines. The _fastcall convention is similar to _stdcall, except the first two available DWORD-sized arguments are passed in %ecx and %edx, respectively. One kludge for this __attribute__ ((regparm(3))), however this isn't entirely right, as it assumes %eax, %ecx and %edx will be used (regparm(2) assumes %eax and %edx). Another kludge is to declare the two fastcall-ed args as local register variables and explicitly assign them to %ecx and %edx, but experimentation showed that gcc would not guard %ecx and %edx against being clobbered. Thus, I came up with a 3rd kludge, which is to use some inline assembly of the form: void *arg1; void *arg2; __asm__("movl %%ecx, %%ecx" : "=c" (arg1)); __asm__("movl %%edx, %%edx" : "=d" (arg2)); This lets gcc know that we're going to reference %ecx and %edx and that it should make an effort not to let it get trampled. This wastes an instruction (movl %reg, %reg is a no-op) but insures proper behavior. It's possible there's a better way to do this though: this is the first time I've used inline assembler in this fashion. The above fixes to ntoskrnl_var.h an subr_ntoskrnl.c make lookaside lists work for the two drivers I have that use them, one of which is an NDIS 5.0 miniport and another which is 5.1.
2003-12-13 07:41:12 +00:00
__stdcall static void ndis_free_string(ndis_unicode_string *);
__stdcall static ndis_status ndis_remove_miniport(ndis_handle *);
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void ndis_termwrap(ndis_handle, void *);
__stdcall static void ndis_get_devprop(ndis_handle, void *, void *,
void *, cm_resource_list *, cm_resource_list *);
Now that I finally have power back, implement a couple more NDIS API routines: NdisUnchainBufferAtBack(), NdisGetFirstBufferFromPacketSafe() and NdisGetFirstBufferFromPacket(). This should bring us a little closer to getting the Intel centrino wireless NIC to work. Note: I have not actually tested these additions since I don't have a driver that calls them, however they're pretty simple, and one of them is taken pretty much directly from the Windows ndis.h header file, so I'm fairly confident they work, but disclaimers apply.
2003-12-22 08:24:32 +00:00
__stdcall static void ndis_firstbuf(ndis_packet *, ndis_buffer **,
void **, uint32_t *, uint32_t *);
__stdcall static void ndis_firstbuf_safe(ndis_packet *, ndis_buffer **,
void **, uint32_t *, uint32_t *, uint32_t);
- Add stubs for Ndis*File() functions - Fix ndis_time(). - Implement NdisGetSystemUpTime(). - Implement RtlCopyUnicodeString() and RtlUnicodeStringToAnsiString(). - In ndis_getstate_80211(), use sc->ndis_link to determine connect status. Submitted by: Brian Feldman <green@freebsd.org>
2003-12-25 00:40:02 +00:00
__stdcall static void ndis_open_file(ndis_status *, ndis_handle *, uint32_t *,
ndis_unicode_string *, ndis_physaddr);
__stdcall static void ndis_map_file(ndis_status *, void **, ndis_handle);
__stdcall static void ndis_unmap_file(ndis_handle);
__stdcall static void ndis_close_file(ndis_handle);
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
__stdcall static u_int8_t ndis_cpu_cnt(void);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void dummy(void);
int
ndis_libinit()
{
if (ndis_inits) {
ndis_inits++;
return(0);
}
mtx_init(&ndis_interlock, "ndislock", MTX_NETWORK_LOCK,
MTX_DEF | MTX_RECURSE | MTX_DUPOK);
ndis_inits++;
return(0);
}
int
ndis_libfini()
{
if (ndis_inits != 1) {
ndis_inits--;
return(0);
}
mtx_destroy(&ndis_interlock);
ndis_inits--;
return(0);
}
/*
* NDIS deals with strings in unicode format, so we have
* do deal with them that way too. For now, we only handle
* conversion between unicode and ASCII since that's all
* that device drivers care about.
*/
int
ndis_ascii_to_unicode(ascii, unicode)
char *ascii;
uint16_t **unicode;
{
uint16_t *ustr;
int i;
if (*unicode == NULL)
*unicode = malloc(strlen(ascii) * 2, M_DEVBUF, M_WAITOK);
if (*unicode == NULL)
return(ENOMEM);
ustr = *unicode;
for (i = 0; i < strlen(ascii); i++) {
*ustr = (uint16_t)ascii[i];
ustr++;
}
return(0);
}
int
ndis_unicode_to_ascii(unicode, ulen, ascii)
uint16_t *unicode;
int ulen;
char **ascii;
{
uint8_t *astr;
int i;
if (*ascii == NULL)
*ascii = malloc(ulen, M_DEVBUF, M_WAITOK);
if (*ascii == NULL)
return(ENOMEM);
astr = *ascii;
for (i = 0; i < ulen; i++) {
*astr = (uint8_t)unicode[i];
astr++;
}
return(0);
}
__stdcall static void
ndis_initwrap(wrapper, drv_obj, path, unused)
ndis_handle wrapper;
ndis_driver_object *drv_obj;
void *path;
void *unused;
{
ndis_driver_object **drv;
drv = wrapper;
*drv = drv_obj;
return;
}
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void
ndis_termwrap(handle, syspec)
ndis_handle handle;
void *syspec;
{
return;
}
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static ndis_status
ndis_register_miniport(handle, characteristics, len)
ndis_handle handle;
ndis_miniport_characteristics *characteristics;
int len;
{
ndis_driver_object *drv;
drv = handle;
bcopy((char *)characteristics, (char *)&drv->ndo_chars,
sizeof(ndis_miniport_characteristics));
return(NDIS_STATUS_SUCCESS);
}
__stdcall static ndis_status
ndis_malloc_withtag(vaddr, len, tag)
void **vaddr;
uint32_t len;
uint32_t tag;
{
void *mem;
mem = malloc(len, M_DEVBUF, M_NOWAIT);
if (mem == NULL)
return(NDIS_STATUS_RESOURCES);
*vaddr = mem;
return(NDIS_STATUS_SUCCESS);
}
__stdcall static ndis_status
ndis_malloc(vaddr, len, flags, highaddr)
void **vaddr;
uint32_t len;
uint32_t flags;
ndis_physaddr highaddr;
{
void *mem;
mem = malloc(len, M_DEVBUF, M_NOWAIT);
if (mem == NULL)
return(NDIS_STATUS_RESOURCES);
*vaddr = mem;
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_free(vaddr, len, flags)
void *vaddr;
uint32_t len;
uint32_t flags;
{
if (len == 0)
return;
free(vaddr, M_DEVBUF);
return;
}
__stdcall static ndis_status
ndis_setattr_ex(adapter_handle, adapter_ctx, hangsecs,
flags, iftype)
ndis_handle adapter_handle;
ndis_handle adapter_ctx;
uint32_t hangsecs;
uint32_t flags;
ndis_interface_type iftype;
{
ndis_miniport_block *block;
/*
* Save the adapter context, we need it for calling
* the driver's internal functions.
*/
block = (ndis_miniport_block *)adapter_handle;
block->nmb_miniportadapterctx = adapter_ctx;
block->nmb_checkforhangsecs = hangsecs;
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_open_cfg(status, cfg, wrapctx)
ndis_status *status;
ndis_handle *cfg;
ndis_handle wrapctx;
{
*cfg = wrapctx;
*status = NDIS_STATUS_SUCCESS;
return;
}
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void
ndis_open_cfgbyname(status, cfg, subkey, subhandle)
ndis_status *status;
ndis_handle cfg;
ndis_unicode_string *subkey;
ndis_handle *subhandle;
{
*subhandle = cfg;
*status = NDIS_STATUS_SUCCESS;
return;
}
__stdcall static void
ndis_open_cfgbyidx(status, cfg, idx, subkey, subhandle)
ndis_status *status;
ndis_handle cfg;
uint32_t idx;
ndis_unicode_string *subkey;
ndis_handle *subhandle;
{
*status = NDIS_STATUS_FAILURE;
return;
}
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
static ndis_status
ndis_encode_parm(block, oid, type, parm)
ndis_miniport_block *block;
struct sysctl_oid *oid;
ndis_parm_type type;
ndis_config_parm **parm;
{
uint16_t *unicode;
ndis_unicode_string *ustr;
unicode = (uint16_t *)&block->nmb_dummybuf;
switch(type) {
case ndis_parm_string:
ndis_ascii_to_unicode((char *)oid->oid_arg1, &unicode);
(*parm)->ncp_type = ndis_parm_string;
ustr = &(*parm)->ncp_parmdata.ncp_stringdata;
ustr->nus_len = strlen((char *)oid->oid_arg1) * 2;
ustr->nus_buf = unicode;
break;
case ndis_parm_int:
(*parm)->ncp_type = ndis_parm_int;
(*parm)->ncp_parmdata.ncp_intdata =
strtol((char *)oid->oid_arg1, NULL, 10);
break;
case ndis_parm_hexint:
(*parm)->ncp_type = ndis_parm_hexint;
(*parm)->ncp_parmdata.ncp_intdata =
strtoul((char *)oid->oid_arg1, NULL, 16);
break;
default:
return(NDIS_STATUS_FAILURE);
break;
}
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_read_cfg(status, parm, cfg, key, type)
ndis_status *status;
ndis_config_parm **parm;
ndis_handle cfg;
ndis_unicode_string *key;
ndis_parm_type type;
{
char *keystr = NULL;
uint16_t *unicode;
ndis_miniport_block *block;
struct ndis_softc *sc;
struct sysctl_oid *oidp;
struct sysctl_ctx_entry *e;
block = (ndis_miniport_block *)cfg;
sc = (struct ndis_softc *)block->nmb_ifp;
ndis_unicode_to_ascii(key->nus_buf, key->nus_len, &keystr);
*parm = &block->nmb_replyparm;
bzero((char *)&block->nmb_replyparm, sizeof(ndis_config_parm));
unicode = (uint16_t *)&block->nmb_dummybuf;
/*
* See if registry key is already in a list of known keys
* included with the driver.
*/
TAILQ_FOREACH(e, &sc->ndis_ctx, link) {
oidp = e->entry;
if (strcmp(oidp->oid_name, keystr) == 0) {
Correct the behavior of ndis_adjust_buflen(): the NDIS spec says it's an error to set the buffer bytecount to anything larger than the buffer's original allocation size, but anything less than that is ok. Also, in ndis_ptom(), use the same logic: if the bytecount is larger than the allocation size, consider the bytecount invalid and the allocation size as the packet fragment length (m_len) instead of the bytecount. This corrects a consistency problem between the Broadcom wireless driver and some of the ethernet drivers I've tested: the ethernet drivers all report the packet frag sizes in buf->nb_bytecount, but the Broadcom wireless driver reports them in buf->nb_size. This seems like a bug to me, but it clearly must work in Windows, so we have to deal with it here too.
2003-12-12 08:54:48 +00:00
if (strcmp((char *)oidp->oid_arg1, "UNSET") == 0) {
free(keystr, M_DEVBUF);
*status = NDIS_STATUS_FAILURE;
return;
}
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
*status = ndis_encode_parm(block, oidp, type, parm);
free(keystr, M_DEVBUF);
return;
}
}
/*
* If the key didn't match, add it to the list of dynamically
* created ones. Sometimes, drivers refer to registry keys
* that aren't documented in their .INF files. These keys
* are supposed to be created by some sort of utility or
* control panel snap-in that comes with the driver software.
* Sometimes it's useful to be able to manipulate these.
* If the driver requests the key in the form of a string,
* make its default value an empty string, otherwise default
* it to "0".
*/
if (type == ndis_parm_int || type == ndis_parm_hexint)
Correct the behavior of ndis_adjust_buflen(): the NDIS spec says it's an error to set the buffer bytecount to anything larger than the buffer's original allocation size, but anything less than that is ok. Also, in ndis_ptom(), use the same logic: if the bytecount is larger than the allocation size, consider the bytecount invalid and the allocation size as the packet fragment length (m_len) instead of the bytecount. This corrects a consistency problem between the Broadcom wireless driver and some of the ethernet drivers I've tested: the ethernet drivers all report the packet frag sizes in buf->nb_bytecount, but the Broadcom wireless driver reports them in buf->nb_size. This seems like a bug to me, but it clearly must work in Windows, so we have to deal with it here too.
2003-12-12 08:54:48 +00:00
ndis_add_sysctl(sc, keystr, "(dynamic integer key)",
"UNSET", CTLFLAG_RW);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
else
Correct the behavior of ndis_adjust_buflen(): the NDIS spec says it's an error to set the buffer bytecount to anything larger than the buffer's original allocation size, but anything less than that is ok. Also, in ndis_ptom(), use the same logic: if the bytecount is larger than the allocation size, consider the bytecount invalid and the allocation size as the packet fragment length (m_len) instead of the bytecount. This corrects a consistency problem between the Broadcom wireless driver and some of the ethernet drivers I've tested: the ethernet drivers all report the packet frag sizes in buf->nb_bytecount, but the Broadcom wireless driver reports them in buf->nb_size. This seems like a bug to me, but it clearly must work in Windows, so we have to deal with it here too.
2003-12-12 08:54:48 +00:00
ndis_add_sysctl(sc, keystr, "(dynamic string key)",
"UNSET", CTLFLAG_RW);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
free(keystr, M_DEVBUF);
*status = NDIS_STATUS_FAILURE;
return;
}
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
static ndis_status
ndis_decode_parm(block, parm, val)
ndis_miniport_block *block;
ndis_config_parm *parm;
char *val;
{
uint16_t *unicode;
ndis_unicode_string *ustr;
unicode = (uint16_t *)&block->nmb_dummybuf;
switch(parm->ncp_type) {
case ndis_parm_string:
ustr = &parm->ncp_parmdata.ncp_stringdata;
ndis_unicode_to_ascii(ustr->nus_buf, ustr->nus_len, &val);
break;
case ndis_parm_int:
sprintf(val, "%ul", parm->ncp_parmdata.ncp_intdata);
break;
case ndis_parm_hexint:
sprintf(val, "%xu", parm->ncp_parmdata.ncp_intdata);
break;
default:
return(NDIS_STATUS_FAILURE);
break;
}
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_write_cfg(status, cfg, key, parm)
ndis_status *status;
ndis_handle cfg;
ndis_unicode_string *key;
ndis_config_parm *parm;
{
char *keystr = NULL;
ndis_miniport_block *block;
struct ndis_softc *sc;
struct sysctl_oid *oidp;
struct sysctl_ctx_entry *e;
char val[256];
block = (ndis_miniport_block *)cfg;
sc = (struct ndis_softc *)block->nmb_ifp;
ndis_unicode_to_ascii(key->nus_buf, key->nus_len, &keystr);
/* Decode the parameter into a string. */
*status = ndis_decode_parm(block, parm, val);
if (*status != NDIS_STATUS_SUCCESS) {
free(keystr, M_DEVBUF);
return;
}
/* See if the key already exists. */
TAILQ_FOREACH(e, &sc->ndis_ctx, link) {
oidp = e->entry;
if (strcmp(oidp->oid_name, keystr) == 0) {
/* Found it, set the value. */
strcpy((char *)oidp->oid_arg1, val);
free(keystr, M_DEVBUF);
return;
}
}
/* Not found, add a new key with the specified value. */
ndis_add_sysctl(sc, keystr, "(dynamically set key)",
val, CTLFLAG_RW);
free(keystr, M_DEVBUF);
*status = NDIS_STATUS_SUCCESS;
return;
}
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void
ndis_close_cfg(cfg)
ndis_handle cfg;
{
return;
}
__stdcall static void
ndis_create_lock(lock)
ndis_spin_lock *lock;
{
struct mtx *mtx;
mtx = malloc(sizeof(struct mtx), M_DEVBUF, M_NOWAIT|M_ZERO);
if (mtx == NULL)
return;
Big round of updates: - Make ndis_get_info()/ndis_set_info() sleep on the setdone/getdone routines if they get back NDIS_STATUS_PENDING. - Add a bunch of net80211 support so that 802.11 cards can be twiddled with ifconfig. This still needs more work and is not guaranteed to work for everyone. It works on my 802.11b/g card anyway. The problem here is Microsoft doesn't provide a good way to a) learn all the rates that a card supports (if it has more than 8, you're kinda hosed) and b) doesn't provide a good way to distinguish between 802.11b, 802.11b/g an 802.11a/b/g cards, so you sort of have to guess. Setting the SSID and switching between infrastructure/adhoc modes should work. WEP still needs to be implemented. I can't find any API for getting/setting the channel other than the registry/sysctl keys.
2003-12-21 00:00:08 +00:00
mtx_init(mtx, "ndislock", "ndis spin lock",
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
MTX_DEF | MTX_RECURSE | MTX_DUPOK);
lock->nsl_spinlock = (ndis_kspin_lock)mtx;
return;
}
__stdcall static void
ndis_destroy_lock(lock)
ndis_spin_lock *lock;
{
struct mtx *ndis_mtx;
ndis_mtx = (struct mtx *)lock->nsl_spinlock;
mtx_destroy(ndis_mtx);
free(ndis_mtx, M_DEVBUF);
return;
}
__stdcall static void
ndis_lock(lock)
ndis_spin_lock *lock;
{
if (lock == NULL)
return;
mtx_lock((struct mtx *)lock->nsl_spinlock);
return;
}
__stdcall static void
ndis_unlock(lock)
ndis_spin_lock *lock;
{
if (lock == NULL)
return;
mtx_unlock((struct mtx *)lock->nsl_spinlock);
return;
}
__stdcall static uint32_t
ndis_read_pci(adapter, slot, offset, buf, len)
ndis_handle adapter;
uint32_t slot;
uint32_t offset;
void *buf;
uint32_t len;
{
ndis_miniport_block *block;
int i;
char *dest;
block = (ndis_miniport_block *)adapter;
dest = buf;
if (block == NULL || block->nmb_dev == NULL)
return(0);
for (i = 0; i < len; i++)
dest[i] = pci_read_config(block->nmb_dev, i + offset, 1);
return(len);
}
__stdcall static uint32_t
ndis_write_pci(adapter, slot, offset, buf, len)
ndis_handle adapter;
uint32_t slot;
uint32_t offset;
void *buf;
uint32_t len;
{
ndis_miniport_block *block;
int i;
char *dest;
block = (ndis_miniport_block *)adapter;
dest = buf;
if (block == NULL || block->nmb_dev == NULL)
return(0);
for (i = 0; i < len; i++)
pci_write_config(block->nmb_dev, i + offset, dest[i], 1);
return(len);
}
/*
* The errorlog routine uses a variable argument list, so we
* have to declare it this way.
*/
static void
ndis_syslog(ndis_handle adapter, ndis_error_code code,
uint32_t numerrors, ...)
{
ndis_miniport_block *block;
va_list ap;
int i;
block = (ndis_miniport_block *)adapter;
printf ("NDIS ERROR: %x\n", code);
printf ("NDIS NUMERRORS: %x\n", numerrors);
va_start(ap, numerrors);
for (i = 0; i < numerrors; i++)
printf ("argptr: %p\n", va_arg(ap, void *));
va_end(ap);
return;
}
static void
ndis_map_cb(arg, segs, nseg, error)
void *arg;
bus_dma_segment_t *segs;
int nseg;
int error;
{
struct ndis_map_arg *ctx;
int i;
if (error)
return;
ctx = arg;
for (i = 0; i < nseg; i++) {
ctx->nma_fraglist[i].npu_physaddr.np_quad = segs[i].ds_addr;
ctx->nma_fraglist[i].npu_len = segs[i].ds_len;
}
ctx->nma_cnt = nseg;
return;
}
__stdcall static void
ndis_vtophys_load(adapter, buf, mapreg, writedev, addrarray, arraysize)
ndis_handle adapter;
ndis_buffer *buf;
uint32_t mapreg;
uint8_t writedev;
ndis_paddr_unit *addrarray;
uint32_t *arraysize;
{
ndis_miniport_block *block;
struct ndis_softc *sc;
struct ndis_map_arg nma;
bus_dmamap_t map;
int error;
if (adapter == NULL)
return;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)(block->nmb_ifp);
if (mapreg > sc->ndis_mmapcnt)
return;
map = sc->ndis_mmaps[mapreg];
nma.nma_fraglist = addrarray;
error = bus_dmamap_load(sc->ndis_mtag, map,
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
MDL_VA(buf), buf->nb_bytecount, ndis_map_cb,
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
(void *)&nma, BUS_DMA_NOWAIT);
if (error)
return;
bus_dmamap_sync(sc->ndis_mtag, map,
writedev ? BUS_DMASYNC_PREWRITE : BUS_DMASYNC_PREREAD);
*arraysize = nma.nma_cnt;
return;
}
__stdcall static void
ndis_vtophys_unload(adapter, buf, mapreg)
ndis_handle adapter;
ndis_buffer *buf;
uint32_t mapreg;
{
ndis_miniport_block *block;
struct ndis_softc *sc;
bus_dmamap_t map;
if (adapter == NULL)
return;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)(block->nmb_ifp);
if (mapreg > sc->ndis_mmapcnt)
return;
map = sc->ndis_mmaps[mapreg];
bus_dmamap_sync(sc->ndis_mtag, map,
BUS_DMASYNC_POSTREAD|BUS_DMASYNC_POSTWRITE);
bus_dmamap_unload(sc->ndis_mtag, map);
return;
}
__stdcall static void
ndis_create_timer(timer, handle, func, ctx)
ndis_miniport_timer *timer;
ndis_handle *handle;
ndis_timer_function func;
void *ctx;
{
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
struct ndis_timer_entry *ne = NULL;
ndis_miniport_block *block;
block = (ndis_miniport_block *)handle;
ne = malloc(sizeof(struct ndis_timer_entry), M_DEVBUF, M_NOWAIT);
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
callout_init(&ne->nte_ch, CALLOUT_MPSAFE);
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
TAILQ_INSERT_TAIL(&block->nmb_timerlist, ne, link);
ne->nte_timer = timer;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
timer->nmt_ktimer.nk_header.dh_sigstate = TRUE;
timer->nmt_dpc.nk_deferredctx = &ne->nte_ch;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
timer->nmt_timerfunc = func;
timer->nmt_timerctx = ctx;
return;
}
/*
* The driver's timer callout is __stdcall function, so we need this
* intermediate step.
*/
static void
ndis_timercall(arg)
void *arg;
{
ndis_miniport_timer *timer;
__stdcall ndis_timer_function timerfunc;
timer = arg;
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
timer->nmt_ktimer.nk_header.dh_sigstate = FALSE;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
timerfunc = timer->nmt_timerfunc;
timerfunc(NULL, timer->nmt_timerctx, NULL, NULL);
return;
}
/*
* Windows specifies timeouts in milliseconds. We specify timeouts
* in hz. Trying to compute a tenth of a second based on hz is tricky.
* so we approximate. Note that we abuse the dpc portion of the
* miniport timer structure to hold the UNIX callout handle.
*/
__stdcall static void
ndis_set_timer(timer, msecs)
ndis_miniport_timer *timer;
uint32_t msecs;
{
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
struct callout *ch;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
struct timeval tv;
tv.tv_sec = 0;
tv.tv_usec = msecs * 1000;
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
ch = timer->nmt_dpc.nk_deferredctx;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
timer->nmt_dpc.nk_sysarg2 = ndis_timercall;
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
timer->nmt_ktimer.nk_header.dh_sigstate = TRUE;
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
callout_reset(ch, tvtohz(&tv), timer->nmt_dpc.nk_sysarg2, timer);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
return;
}
static void
ndis_tick(arg)
void *arg;
{
ndis_miniport_timer *timer;
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
struct callout *ch;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall ndis_timer_function timerfunc;
struct timeval tv;
timer = arg;
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
timer->nmt_ktimer.nk_header.dh_sigstate = FALSE;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
timerfunc = timer->nmt_timerfunc;
timerfunc(NULL, timer->nmt_timerctx, NULL, NULL);
/* Automatically reload timer. */
tv.tv_sec = 0;
tv.tv_usec = timer->nmt_ktimer.nk_period * 1000;
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
ch = timer->nmt_dpc.nk_deferredctx;
timer->nmt_ktimer.nk_header.dh_sigstate = TRUE;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
timer->nmt_dpc.nk_sysarg2 = ndis_tick;
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
callout_reset(ch, tvtohz(&tv), timer->nmt_dpc.nk_sysarg2, timer);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
return;
}
__stdcall static void
ndis_set_periodic_timer(timer, msecs)
ndis_miniport_timer *timer;
uint32_t msecs;
{
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
struct callout *ch;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
struct timeval tv;
tv.tv_sec = 0;
tv.tv_usec = msecs * 1000;
timer->nmt_ktimer.nk_period = msecs;
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
ch = timer->nmt_dpc.nk_deferredctx;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
timer->nmt_dpc.nk_sysarg2 = ndis_tick;
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
timer->nmt_ktimer.nk_header.dh_sigstate = TRUE;
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
callout_reset(ch, tvtohz(&tv), timer->nmt_dpc.nk_sysarg2, timer);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
return;
}
__stdcall static void
ndis_cancel_timer(timer, cancelled)
ndis_miniport_timer *timer;
uint8_t *cancelled;
{
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
struct callout *ch;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
ch = timer->nmt_dpc.nk_deferredctx;
Give the timer API one last overhaul: this time, use the new callout API instead of the old timeout/untimeout mechanism.
2003-12-25 21:34:16 +00:00
callout_stop(ch);
- Fix some compiler warnings in subr_pe.c - Add explicit cardbus attachment in if_ndis.c - Clean up after moving bus_setup_intr() in ndis_attach(). - When setting an ssid, program an empty ssid as a 1-byte string with a single 0 byte. The Microsoft documentation says this is how you're supposed to tell the NIC to attach to 'any' ssid. - Keep trace of callout handles for timers externally from the ndis_miniport_timer structs, and run through and clobber them all after invoking the haltfunc just in case the driver left one running. (We need to make sure all timers are cancelled on driver unload.) - Handle the 'cancelled' argument in ndis_cancel_timer() correctly.
2003-12-24 21:21:18 +00:00
*cancelled = timer->nmt_ktimer.nk_header.dh_sigstate;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
return;
}
__stdcall static void
ndis_query_resources(status, adapter, list, buflen)
ndis_status *status;
ndis_handle adapter;
ndis_resource_list *list;
uint32_t *buflen;
{
ndis_miniport_block *block;
struct ndis_softc *sc;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)block->nmb_ifp;
*buflen = sizeof(ndis_resource_list) +
(sizeof(cm_partial_resource_desc) * (sc->ndis_rescnt - 1));
bcopy((char *)block->nmb_rlist, (char *)list, *buflen);
*status = NDIS_STATUS_SUCCESS;
return;
}
__stdcall static ndis_status
ndis_register_ioport(offset, adapter, port, numports)
void **offset;
ndis_handle adapter;
uint32_t port;
uint32_t numports;
{
struct ndis_miniport_block *block;
struct ndis_softc *sc;
if (adapter == NULL)
return(NDIS_STATUS_FAILURE);
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)(block->nmb_ifp);
if (sc->ndis_res_io == NULL)
return(NDIS_STATUS_FAILURE);
if (rman_get_size(sc->ndis_res_io) != numports)
return(NDIS_STATUS_INVALID_LENGTH);
*offset = (void *)rman_get_start(sc->ndis_res_io);
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_deregister_ioport(adapter, port, numports, offset)
ndis_handle adapter;
uint32_t port;
uint32_t numports;
void *offset;
{
return;
}
__stdcall static void
ndis_read_netaddr(status, addr, addrlen, adapter)
ndis_status *status;
void **addr;
uint32_t *addrlen;
ndis_handle adapter;
{
struct ndis_softc *sc;
ndis_miniport_block *block;
uint8_t empty[] = { 0, 0, 0, 0, 0, 0 };
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)block->nmb_ifp;
if (bcmp(sc->arpcom.ac_enaddr, empty, ETHER_ADDR_LEN) == 0)
*status = NDIS_STATUS_FAILURE;
else {
*addr = sc->arpcom.ac_enaddr;
*addrlen = ETHER_ADDR_LEN;
*status = NDIS_STATUS_SUCCESS;
}
return;
}
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
__stdcall static ndis_status
ndis_mapreg_cnt(bustype, cnt)
uint32_t bustype;
uint32_t *cnt;
{
*cnt = 64;
return(NDIS_STATUS_SUCCESS);
}
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static ndis_status
ndis_alloc_mapreg(adapter, dmachannel, dmasize, physmapneeded, maxmap)
ndis_handle adapter;
uint32_t dmachannel;
uint8_t dmasize;
uint32_t physmapneeded;
uint32_t maxmap;
{
struct ndis_softc *sc;
ndis_miniport_block *block;
int error, i, nseg = NDIS_MAXSEG;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)block->nmb_ifp;
sc->ndis_mmaps = malloc(sizeof(bus_dmamap_t) * physmapneeded,
M_DEVBUF, M_NOWAIT|M_ZERO);
if (sc->ndis_mmaps == NULL)
return(NDIS_STATUS_RESOURCES);
error = bus_dma_tag_create(sc->ndis_parent_tag, ETHER_ALIGN, 0,
BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL,
NULL, maxmap * nseg, nseg, maxmap, BUS_DMA_ALLOCNOW,
NULL, NULL, &sc->ndis_mtag);
if (error) {
free(sc->ndis_mmaps, M_DEVBUF);
return(NDIS_STATUS_RESOURCES);
}
for (i = 0; i < physmapneeded; i++)
bus_dmamap_create(sc->ndis_mtag, 0, &sc->ndis_mmaps[i]);
sc->ndis_mmapcnt = physmapneeded;
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_free_mapreg(adapter)
ndis_handle adapter;
{
struct ndis_softc *sc;
ndis_miniport_block *block;
int i;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)block->nmb_ifp;
for (i = 0; i < sc->ndis_mmapcnt; i++)
bus_dmamap_destroy(sc->ndis_mtag, sc->ndis_mmaps[i]);
free(sc->ndis_mmaps, M_DEVBUF);
bus_dma_tag_destroy(sc->ndis_mtag);
return;
}
static void
ndis_mapshared_cb(arg, segs, nseg, error)
void *arg;
bus_dma_segment_t *segs;
int nseg;
int error;
{
ndis_physaddr *p;
if (error || nseg > 1)
return;
p = arg;
p->np_quad = segs[0].ds_addr;
return;
}
/*
* This maps to bus_dmamem_alloc().
*/
__stdcall static void
ndis_alloc_sharedmem(adapter, len, cached, vaddr, paddr)
ndis_handle adapter;
uint32_t len;
uint8_t cached;
void **vaddr;
ndis_physaddr *paddr;
{
ndis_miniport_block *block;
struct ndis_softc *sc;
struct ndis_shmem *sh;
int error;
if (adapter == NULL)
return;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)(block->nmb_ifp);
sh = malloc(sizeof(struct ndis_shmem), M_DEVBUF, M_NOWAIT|M_ZERO);
if (sh == NULL)
return;
error = bus_dma_tag_create(sc->ndis_parent_tag, 64,
0, BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL,
NULL, len, 1, len, BUS_DMA_ALLOCNOW, NULL, NULL,
&sh->ndis_stag);
if (error) {
free(sh, M_DEVBUF);
return;
}
error = bus_dmamem_alloc(sh->ndis_stag, vaddr,
BUS_DMA_NOWAIT | BUS_DMA_ZERO, &sh->ndis_smap);
if (error) {
bus_dma_tag_destroy(sh->ndis_stag);
free(sh, M_DEVBUF);
return;
}
error = bus_dmamap_load(sh->ndis_stag, sh->ndis_smap, *vaddr,
len, ndis_mapshared_cb, (void *)paddr, BUS_DMA_NOWAIT);
if (error) {
bus_dmamem_free(sh->ndis_stag, *vaddr, sh->ndis_smap);
bus_dma_tag_destroy(sh->ndis_stag);
free(sh, M_DEVBUF);
return;
}
sh->ndis_saddr = *vaddr;
sh->ndis_next = sc->ndis_shlist;
sc->ndis_shlist = sh;
return;
}
__stdcall static void
ndis_alloc_sharedmem_async(adapter, len, cached, ctx)
ndis_handle adapter;
uint32_t len;
uint8_t cached;
void *ctx;
{
ndis_miniport_block *block;
struct ndis_softc *sc;
void *vaddr;
ndis_physaddr paddr;
__stdcall ndis_allocdone_handler donefunc;
if (adapter == NULL)
return;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)(block->nmb_ifp);
donefunc = sc->ndis_chars.nmc_allocate_complete_func;
ndis_alloc_sharedmem(adapter, len, cached, &vaddr, &paddr);
donefunc(adapter, vaddr, &paddr, len, ctx);
return;
}
__stdcall static void
ndis_free_sharedmem(adapter, len, cached, vaddr, paddr)
ndis_handle adapter;
uint32_t len;
uint8_t cached;
void *vaddr;
ndis_physaddr paddr;
{
ndis_miniport_block *block;
struct ndis_softc *sc;
struct ndis_shmem *sh, *prev;
if (vaddr == NULL || adapter == NULL)
return;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)(block->nmb_ifp);
sh = prev = sc->ndis_shlist;
while (sh) {
if (sh->ndis_saddr == vaddr)
break;
prev = sh;
sh = sh->ndis_next;
}
bus_dmamap_unload(sh->ndis_stag, sh->ndis_smap);
bus_dmamem_free(sh->ndis_stag, vaddr, sh->ndis_smap);
bus_dma_tag_destroy(sh->ndis_stag);
if (sh == sc->ndis_shlist)
sc->ndis_shlist = sh->ndis_next;
else
prev->ndis_next = sh->ndis_next;
free(sh, M_DEVBUF);
return;
}
__stdcall static ndis_status
ndis_map_iospace(vaddr, adapter, paddr, len)
void **vaddr;
ndis_handle adapter;
ndis_physaddr paddr;
uint32_t len;
{
ndis_miniport_block *block;
struct ndis_softc *sc;
if (adapter == NULL)
return(NDIS_STATUS_FAILURE);
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)(block->nmb_ifp);
if (sc->ndis_res_mem == NULL)
return(NDIS_STATUS_FAILURE);
*vaddr = (void *)rman_get_virtual(sc->ndis_res_mem);
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_unmap_iospace(adapter, vaddr, len)
ndis_handle adapter;
void *vaddr;
uint32_t len;
{
return;
}
__stdcall static uint32_t
ndis_cachefill(void)
{
return(128);
}
__stdcall static uint32_t
ndis_dma_align(handle)
ndis_handle handle;
{
return(128);
}
/*
* NDIS has two methods for dealing with NICs that support DMA.
* One is to just pass packets to the driver and let it call
* NdisMStartBufferPhysicalMapping() to map each buffer in the packet
* all by itself, and the other is to let the NDIS library handle the
* buffer mapping internally, and hand the driver an already populated
* scatter/gather fragment list. If the driver calls
* NdisMInitializeScatterGatherDma(), it wants to use the latter
* method.
*/
__stdcall static ndis_status
ndis_init_sc_dma(adapter, is64, maxphysmap)
ndis_handle adapter;
uint8_t is64;
uint32_t maxphysmap;
{
struct ndis_softc *sc;
ndis_miniport_block *block;
int error;
if (adapter == NULL)
return(NDIS_STATUS_FAILURE);
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)block->nmb_ifp;
/* Don't do this twice. */
if (sc->ndis_sc == 1)
return(NDIS_STATUS_SUCCESS);
error = bus_dma_tag_create(sc->ndis_parent_tag, ETHER_ALIGN, 0,
BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL, NULL,
MCLBYTES * NDIS_MAXSEG, NDIS_MAXSEG, MCLBYTES, BUS_DMA_ALLOCNOW,
NULL, NULL, &sc->ndis_ttag);
sc->ndis_sc = 1;
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_alloc_packetpool(status, pool, descnum, protrsvdlen)
ndis_status *status;
ndis_handle *pool;
uint32_t descnum;
uint32_t protrsvdlen;
{
ndis_packet *cur;
int i;
*pool = malloc(sizeof(ndis_packet) * (descnum + 1),
M_DEVBUF, M_NOWAIT|M_ZERO);
if (pool == NULL) {
*status = NDIS_STATUS_RESOURCES;
return;
}
cur = (ndis_packet *)*pool;
cur->np_private.npp_flags = 0x1; /* mark the head of the list */
for (i = 0; i < descnum; i++) {
cur->np_private.npp_head = (ndis_handle)(cur + 1);
cur++;
}
*status = NDIS_STATUS_SUCCESS;
return;
}
__stdcall static void
ndis_ex_alloc_packetpool(status, pool, descnum, oflowdescnum, protrsvdlen)
ndis_status *status;
ndis_handle *pool;
uint32_t descnum;
uint32_t oflowdescnum;
uint32_t protrsvdlen;
{
return(ndis_alloc_packetpool(status, pool,
descnum + oflowdescnum, protrsvdlen));
}
__stdcall static uint32_t
ndis_packetpool_use(pool)
ndis_handle pool;
{
ndis_packet *head;
head = (ndis_packet *)pool;
return(head->np_private.npp_count);
}
__stdcall static void
ndis_free_packetpool(pool)
ndis_handle pool;
{
free(pool, M_DEVBUF);
return;
}
__stdcall static void
ndis_alloc_packet(status, packet, pool)
ndis_status *status;
ndis_packet **packet;
ndis_handle pool;
{
ndis_packet *head, *pkt;
head = (ndis_packet *)pool;
if (head->np_private.npp_flags != 0x1) {
*status = NDIS_STATUS_FAILURE;
return;
}
pkt = (ndis_packet *)head->np_private.npp_head;
if (pkt == NULL) {
*status = NDIS_STATUS_RESOURCES;
return;
}
head->np_private.npp_head = pkt->np_private.npp_head;
pkt->np_private.npp_head = pkt->np_private.npp_tail = NULL;
/* Save pointer to the pool. */
pkt->np_private.npp_pool = head;
/* Set the oob offset pointer. Lots of things expect this. */
pkt->np_private.npp_packetooboffset =
offsetof(ndis_packet, np_oob);
*packet = pkt;
head->np_private.npp_count++;
*status = NDIS_STATUS_SUCCESS;
return;
}
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
__stdcall static void
ndis_release_packet(packet)
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
ndis_packet *packet;
{
ndis_packet *head;
if (packet == NULL || packet->np_private.npp_pool == NULL)
return;
head = packet->np_private.npp_pool;
if (head->np_private.npp_flags != 0x1)
return;
packet->np_private.npp_head = head->np_private.npp_head;
head->np_private.npp_head = (ndis_buffer *)packet;
head->np_private.npp_count--;
return;
}
__stdcall static void
ndis_unchain_headbuf(packet, buf)
ndis_packet *packet;
ndis_buffer **buf;
{
ndis_packet_private *priv;
if (packet == NULL || buf == NULL)
return;
priv = &packet->np_private;
priv->npp_validcounts = FALSE;
if (priv->npp_head == priv->npp_tail) {
*buf = priv->npp_head;
priv->npp_head = priv->npp_tail = NULL;
} else {
*buf = priv->npp_head;
priv->npp_head = (*buf)->nb_next;
}
return;
}
Now that I finally have power back, implement a couple more NDIS API routines: NdisUnchainBufferAtBack(), NdisGetFirstBufferFromPacketSafe() and NdisGetFirstBufferFromPacket(). This should bring us a little closer to getting the Intel centrino wireless NIC to work. Note: I have not actually tested these additions since I don't have a driver that calls them, however they're pretty simple, and one of them is taken pretty much directly from the Windows ndis.h header file, so I'm fairly confident they work, but disclaimers apply.
2003-12-22 08:24:32 +00:00
__stdcall static void
ndis_unchain_tailbuf(packet, buf)
ndis_packet *packet;
ndis_buffer **buf;
{
ndis_packet_private *priv;
ndis_buffer *tmp;
if (packet == NULL || buf == NULL)
return;
priv = &packet->np_private;
priv->npp_validcounts = FALSE;
if (priv->npp_head == priv->npp_tail) {
*buf = priv->npp_head;
priv->npp_head = priv->npp_tail = NULL;
} else {
*buf = priv->npp_tail;
tmp = priv->npp_head;
while (tmp->nb_next != priv->npp_tail)
tmp = tmp->nb_next;
priv->npp_tail = tmp;
tmp->nb_next = NULL;
}
return;
}
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
/*
* The NDIS "buffer" manipulation functions are somewhat misnamed.
* They don't really allocate buffers: they allocate buffer mappings.
* The idea is you reserve a chunk of DMA-able memory using
* NdisMAllocateSharedMemory() and then use NdisAllocateBuffer()
* to obtain the virtual address of the DMA-able region.
* ndis_alloc_bufpool() is analagous to bus_dma_tag_create().
*/
__stdcall static void
ndis_alloc_bufpool(status, pool, descnum)
ndis_status *status;
ndis_handle *pool;
uint32_t descnum;
{
ndis_buffer *cur;
int i;
*pool = malloc(sizeof(ndis_buffer) * (descnum + 1),
M_DEVBUF, M_NOWAIT|M_ZERO);
if (pool == NULL) {
*status = NDIS_STATUS_RESOURCES;
return;
}
cur = (ndis_buffer *)*pool;
cur->nb_flags = 0x1; /* mark the head of the list */
for (i = 0; i < descnum; i++) {
cur->nb_next = cur + 1;
cur++;
}
*status = NDIS_STATUS_SUCCESS;
return;
}
__stdcall static void
ndis_free_bufpool(pool)
ndis_handle pool;
{
free(pool, M_DEVBUF);
return;
}
/*
* This maps to a bus_dmamap_create() and bus_dmamap_load().
*/
__stdcall static void
ndis_alloc_buf(status, buffer, pool, vaddr, len)
ndis_status *status;
ndis_buffer **buffer;
ndis_handle pool;
void *vaddr;
uint32_t len;
{
ndis_buffer *head, *buf;
head = (ndis_buffer *)pool;
if (head->nb_flags != 0x1) {
*status = NDIS_STATUS_FAILURE;
return;
}
buf = head->nb_next;
if (buf == NULL) {
*status = NDIS_STATUS_RESOURCES;
return;
}
head->nb_next = buf->nb_next;
/* Save pointer to the pool. */
buf->nb_process = head;
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
MDL_INIT(buf, vaddr, len);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
*buffer = buf;
*status = NDIS_STATUS_SUCCESS;
return;
}
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
__stdcall static void
ndis_release_buf(buf)
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
ndis_buffer *buf;
{
ndis_buffer *head;
if (buf == NULL || buf->nb_process == NULL)
return;
head = buf->nb_process;
if (head->nb_flags != 0x1)
return;
buf->nb_next = head->nb_next;
head->nb_next = buf;
return;
}
Some minor touchups: In NdisQueryBuffer() and NdisQueryBufferSafe(), the vaddr argument is optional, so test it before trying to dereference it. Also correct NdisGetFirstBufferFromPacket()/NdisGetFirstBufferFromPacketSafe(): we need to use nb_mappedsystemva from the buffer, not nb_systemva.
2003-12-22 10:22:57 +00:00
/*
* Get the virtual address and length of a buffer.
* Note: the vaddr argument is optional.
*/
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void
ndis_query_buf(buf, vaddr, len)
ndis_buffer *buf;
void **vaddr;
uint32_t *len;
{
Some minor touchups: In NdisQueryBuffer() and NdisQueryBufferSafe(), the vaddr argument is optional, so test it before trying to dereference it. Also correct NdisGetFirstBufferFromPacket()/NdisGetFirstBufferFromPacketSafe(): we need to use nb_mappedsystemva from the buffer, not nb_systemva.
2003-12-22 10:22:57 +00:00
if (vaddr != NULL)
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
*vaddr = MDL_VA(buf);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
*len = buf->nb_bytecount;
return;
}
/* Same as above -- we don't care about the priority. */
__stdcall static void
ndis_query_buf_safe(buf, vaddr, len, prio)
ndis_buffer *buf;
void **vaddr;
uint32_t *len;
uint32_t prio;
{
Some minor touchups: In NdisQueryBuffer() and NdisQueryBufferSafe(), the vaddr argument is optional, so test it before trying to dereference it. Also correct NdisGetFirstBufferFromPacket()/NdisGetFirstBufferFromPacketSafe(): we need to use nb_mappedsystemva from the buffer, not nb_systemva.
2003-12-22 10:22:57 +00:00
if (vaddr != NULL)
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
*vaddr = MDL_VA(buf);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
*len = buf->nb_bytecount;
return;
}
__stdcall static void
ndis_adjust_buflen(buf, len)
ndis_buffer *buf;
int len;
{
buf->nb_bytecount = len;
return;
}
__stdcall static uint32_t
ndis_interlock_inc(addend)
uint32_t *addend;
{
mtx_lock(&ndis_interlock);
*addend++;
mtx_unlock(&ndis_interlock);
return(*addend);
}
__stdcall static uint32_t
ndis_interlock_dec(addend)
uint32_t *addend;
{
mtx_lock(&ndis_interlock);
*addend--;
mtx_unlock(&ndis_interlock);
return(*addend);
}
__stdcall static void
ndis_init_event(event)
ndis_event *event;
{
event->ne_event.nk_header.dh_sigstate = FALSE;
return;
}
__stdcall static void
ndis_set_event(event)
ndis_event *event;
{
event->ne_event.nk_header.dh_sigstate = TRUE;
wakeup(event);
return;
}
__stdcall static void
ndis_reset_event(event)
ndis_event *event;
{
event->ne_event.nk_header.dh_sigstate = FALSE;
wakeup(event);
return;
}
__stdcall static uint8_t
ndis_wait_event(event, msecs)
ndis_event *event;
uint32_t msecs;
{
int error;
struct timeval tv;
if (event->ne_event.nk_header.dh_sigstate == TRUE)
return(TRUE);
tv.tv_sec = 0;
tv.tv_usec = msecs * 1000;
error = tsleep(event, PPAUSE|PCATCH, "ndis", tvtohz(&tv));
return(event->ne_event.nk_header.dh_sigstate);
}
__stdcall static ndis_status
ndis_unicode2ansi(dstr, sstr)
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
ndis_ansi_string *dstr;
ndis_unicode_string *sstr;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
{
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
if (dstr == NULL || sstr == NULL)
return(NDIS_STATUS_FAILURE);
if (ndis_unicode_to_ascii(sstr->nus_buf,
sstr->nus_len, &dstr->nas_buf))
return(NDIS_STATUS_FAILURE);
dstr->nas_len = dstr->nas_maxlen = strlen(dstr->nas_buf);
return (NDIS_STATUS_SUCCESS);
}
__stdcall static ndis_status
ndis_ansi2unicode(dstr, sstr)
ndis_unicode_string *dstr;
ndis_ansi_string *sstr;
{
char *str;
if (dstr == NULL || sstr == NULL)
return(NDIS_STATUS_FAILURE);
str = malloc(sstr->nas_len + 1, M_DEVBUF, M_NOWAIT);
if (str == NULL)
return(NDIS_STATUS_FAILURE);
strncpy(str, sstr->nas_buf, sstr->nas_len);
*(str + sstr->nas_len) = '\0';
if (ndis_ascii_to_unicode(str, &dstr->nus_buf)) {
free(str, M_DEVBUF);
return(NDIS_STATUS_FAILURE);
}
dstr->nus_len = dstr->nus_maxlen = sstr->nas_len * 2;
free(str, M_DEVBUF);
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
return (NDIS_STATUS_SUCCESS);
}
__stdcall static ndis_status
ndis_assign_pcirsrc(adapter, slot, list)
ndis_handle adapter;
uint32_t slot;
ndis_resource_list **list;
{
ndis_miniport_block *block;
if (adapter == NULL || list == NULL)
return (NDIS_STATUS_FAILURE);
block = (ndis_miniport_block *)adapter;
*list = block->nmb_rlist;
printf ("assign PCI resources...\n");
return (NDIS_STATUS_SUCCESS);
}
__stdcall static ndis_status
ndis_register_intr(intr, adapter, ivec, ilevel, reqisr, shared, imode)
ndis_miniport_interrupt *intr;
ndis_handle adapter;
uint32_t ivec;
uint32_t ilevel;
uint8_t reqisr;
uint8_t shared;
ndis_interrupt_mode imode;
{
return(NDIS_STATUS_SUCCESS);
}
__stdcall static void
ndis_deregister_intr(intr)
ndis_miniport_interrupt *intr;
{
return;
}
__stdcall static void
ndis_register_shutdown(adapter, shutdownctx, shutdownfunc)
ndis_handle adapter;
void *shutdownctx;
ndis_shutdown_handler shutdownfunc;
{
ndis_miniport_block *block;
ndis_miniport_characteristics *chars;
struct ndis_softc *sc;
if (adapter == NULL)
return;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)block->nmb_ifp;
chars = &sc->ndis_chars;
chars->nmc_shutdown_handler = shutdownfunc;
chars->nmc_rsvd0 = shutdownctx;
return;
}
__stdcall static void
ndis_deregister_shutdown(adapter)
ndis_handle adapter;
{
ndis_miniport_block *block;
ndis_miniport_characteristics *chars;
struct ndis_softc *sc;
if (adapter == NULL)
return;
block = (ndis_miniport_block *)adapter;
sc = (struct ndis_softc *)block->nmb_ifp;
chars = &sc->ndis_chars;
chars->nmc_shutdown_handler = NULL;
chars->nmc_rsvd0 = NULL;
return;
}
__stdcall static uint32_t
ndis_numpages(buf)
ndis_buffer *buf;
{
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
if (buf == NULL)
return(0);
Correct the implementation of NDIS_BUFFER_TO_SPAN_PAGES().
2003-12-13 09:07:35 +00:00
if (buf->nb_bytecount == 0)
return(1);
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
return(SPAN_PAGES(MDL_VA(buf), buf->nb_bytecount));
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
}
Implement NdisGetBufferPhysicalArraySize(), which apparently is a synonym for NDIS_BUFFER_TO_SPAN_PAGES().
2003-12-16 18:56:33 +00:00
__stdcall static void
ndis_buf_physpages(buf, pages)
ndis_buffer *buf;
uint32_t *pages;
{
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
if (buf == NULL)
return;
Implement NdisGetBufferPhysicalArraySize(), which apparently is a synonym for NDIS_BUFFER_TO_SPAN_PAGES().
2003-12-16 18:56:33 +00:00
*pages = ndis_numpages(buf);
return;
}
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void
ndis_query_bufoffset(buf, off, len)
ndis_buffer *buf;
uint32_t *off;
uint32_t *len;
{
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
if (buf == NULL)
return;
*off = buf->nb_byteoffset;
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
*len = buf->nb_bytecount;
return;
}
__stdcall static void
ndis_sleep(usecs)
uint32_t usecs;
{
struct timeval tv;
uint32_t dummy;
tv.tv_sec = 0;
tv.tv_usec = usecs;
tsleep(&dummy, PPAUSE|PCATCH, "ndis", tvtohz(&tv));
return;
}
__stdcall static uint32_t
ndis_read_pccard_amem(handle, offset, buf, len)
ndis_handle handle;
uint32_t offset;
void *buf;
uint32_t len;
{
struct ndis_softc *sc;
ndis_miniport_block *block;
bus_space_handle_t bh;
bus_space_tag_t bt;
char *dest;
int i;
if (handle == NULL)
return(0);
block = (ndis_miniport_block *)handle;
sc = (struct ndis_softc *)block->nmb_ifp;
dest = buf;
bh = rman_get_bushandle(sc->ndis_res_am);
bt = rman_get_bustag(sc->ndis_res_am);
for (i = 0; i < len; i++)
dest[i] = bus_space_read_1(bt, bh, (offset * 2) + (i * 2));
return(i);
}
__stdcall static uint32_t
ndis_write_pccard_amem(handle, offset, buf, len)
ndis_handle handle;
uint32_t offset;
void *buf;
uint32_t len;
{
struct ndis_softc *sc;
ndis_miniport_block *block;
bus_space_handle_t bh;
bus_space_tag_t bt;
char *src;
int i;
if (handle == NULL)
return(0);
block = (ndis_miniport_block *)handle;
sc = (struct ndis_softc *)block->nmb_ifp;
src = buf;
bh = rman_get_bushandle(sc->ndis_res_am);
bt = rman_get_bustag(sc->ndis_res_am);
for (i = 0; i < len; i++)
bus_space_write_1(bt, bh, (offset * 2) + (i * 2), src[i]);
return(i);
}
__stdcall static ndis_list_entry *
ndis_insert_head(head, entry, lock)
ndis_list_entry *head;
ndis_list_entry *entry;
ndis_spin_lock *lock;
{
ndis_list_entry *flink;
mtx_lock_spin((struct mtx *)lock->nsl_spinlock);
flink = head->nle_flink;
entry->nle_flink = flink;
entry->nle_blink = head;
flink->nle_blink = entry;
head->nle_flink = entry;
mtx_unlock_spin((struct mtx *)lock->nsl_spinlock);
return(flink);
}
__stdcall static ndis_list_entry *
ndis_remove_head(head, lock)
ndis_list_entry *head;
ndis_spin_lock *lock;
{
ndis_list_entry *flink;
ndis_list_entry *entry;
mtx_lock_spin((struct mtx *)lock->nsl_spinlock);
entry = head->nle_flink;
flink = entry->nle_flink;
head->nle_flink = flink;
flink->nle_blink = head;
mtx_unlock_spin((struct mtx *)lock->nsl_spinlock);
return(entry);
}
__stdcall static ndis_list_entry *
ndis_insert_tail(head, entry, lock)
ndis_list_entry *head;
ndis_list_entry *entry;
ndis_spin_lock *lock;
{
ndis_list_entry *blink;
mtx_lock_spin((struct mtx *)lock->nsl_spinlock);
blink = head->nle_blink;
entry->nle_flink = head;
entry->nle_blink = blink;
blink->nle_flink = entry;
head->nle_blink = entry;
mtx_unlock_spin((struct mtx *)lock->nsl_spinlock);
return(blink);
}
__stdcall static uint8_t
ndis_sync_with_intr(intr, syncfunc, syncctx)
ndis_miniport_interrupt *intr;
void *syncfunc;
void *syncctx;
{
__stdcall uint8_t (*sync)(void *);
if (syncfunc == NULL || syncctx == NULL)
return(0);
sync = syncfunc;
return(sync(syncctx));
}
Implement some more NDIS and ntoskrnl API calls: subr_ndis.c: NdisGetCurrentSystemTime() which, according to the Microsoft documentation returns "the number of 100 nanosecond intervals since January 1, 1601." I have no idea what's so special about that epoch or why they chose 100 nanosecond ticks. I don't know the proper offset to convert nanotime() from the UNIX epoch to January 1, 1601, so for now I'm just doing the unit convertion to 100s of nanoseconds. subr_ntoskrnl.c: memcpy(), memset(), ExInterlockedPopEntrySList(), ExInterlockedPushEntrySList(). The latter two are different from InterlockedPopEntrySList() and InterlockedPushEntrySList() in that they accept a spinlock to hold while executing, whereas the non-Ex routines use a lock internal to ntoskrnl. I also modified ExInitializePagedLookasideList() and ExInitializeNPagedLookasideList() to initialize mutex locks within the lookaside structures. It seems that in NDIS 5.0, the lookaside allocate/free routines ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(), which require the use of the per-lookaside spinlock, whereas in NDIS 5.1, the per-lookaside spinlock is deprecated. We need to support both cases. Note that I appear to be doing something wrong with ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(): they don't appear to obtain proper pointers to their arguments, so I'm probably doing something wrong in terms of their calling convention (they're declared to be FASTCALL in Widnows, and I'm not sure what that means for gcc). It happens that in my stub lookaside implementation, they don't need to do any work anyway, so for now I've hacked them to always return NULL, which avoids corrupting the stack. I need to do this right though.
2003-12-12 22:35:13 +00:00
/*
* Return the number of 100 nanosecond intervals since
* January 1, 1601. (?!?!)
*/
__stdcall static void
ndis_time(tval)
uint64_t *tval;
{
struct timespec ts;
- Add stubs for Ndis*File() functions - Fix ndis_time(). - Implement NdisGetSystemUpTime(). - Implement RtlCopyUnicodeString() and RtlUnicodeStringToAnsiString(). - In ndis_getstate_80211(), use sc->ndis_link to determine connect status. Submitted by: Brian Feldman <green@freebsd.org>
2003-12-25 00:40:02 +00:00
Implement some more NDIS and ntoskrnl API calls: subr_ndis.c: NdisGetCurrentSystemTime() which, according to the Microsoft documentation returns "the number of 100 nanosecond intervals since January 1, 1601." I have no idea what's so special about that epoch or why they chose 100 nanosecond ticks. I don't know the proper offset to convert nanotime() from the UNIX epoch to January 1, 1601, so for now I'm just doing the unit convertion to 100s of nanoseconds. subr_ntoskrnl.c: memcpy(), memset(), ExInterlockedPopEntrySList(), ExInterlockedPushEntrySList(). The latter two are different from InterlockedPopEntrySList() and InterlockedPushEntrySList() in that they accept a spinlock to hold while executing, whereas the non-Ex routines use a lock internal to ntoskrnl. I also modified ExInitializePagedLookasideList() and ExInitializeNPagedLookasideList() to initialize mutex locks within the lookaside structures. It seems that in NDIS 5.0, the lookaside allocate/free routines ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(), which require the use of the per-lookaside spinlock, whereas in NDIS 5.1, the per-lookaside spinlock is deprecated. We need to support both cases. Note that I appear to be doing something wrong with ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(): they don't appear to obtain proper pointers to their arguments, so I'm probably doing something wrong in terms of their calling convention (they're declared to be FASTCALL in Widnows, and I'm not sure what that means for gcc). It happens that in my stub lookaside implementation, they don't need to do any work anyway, so for now I've hacked them to always return NULL, which avoids corrupting the stack. I need to do this right though.
2003-12-12 22:35:13 +00:00
nanotime(&ts);
- Add stubs for Ndis*File() functions - Fix ndis_time(). - Implement NdisGetSystemUpTime(). - Implement RtlCopyUnicodeString() and RtlUnicodeStringToAnsiString(). - In ndis_getstate_80211(), use sc->ndis_link to determine connect status. Submitted by: Brian Feldman <green@freebsd.org>
2003-12-25 00:40:02 +00:00
*tval = (uint64_t)ts.tv_nsec / 100 + (uint64_t)ts.tv_sec * 10000000 +
11644473600;
}
/*
* Return the number of milliseconds since the system booted.
*/
__stdcall static void
ndis_uptime(tval)
uint32_t *tval;
{
struct timespec ts;
nanouptime(&ts);
*tval = ts.tv_nsec / 1000000 + ts.tv_sec * 1000;
Implement some more NDIS and ntoskrnl API calls: subr_ndis.c: NdisGetCurrentSystemTime() which, according to the Microsoft documentation returns "the number of 100 nanosecond intervals since January 1, 1601." I have no idea what's so special about that epoch or why they chose 100 nanosecond ticks. I don't know the proper offset to convert nanotime() from the UNIX epoch to January 1, 1601, so for now I'm just doing the unit convertion to 100s of nanoseconds. subr_ntoskrnl.c: memcpy(), memset(), ExInterlockedPopEntrySList(), ExInterlockedPushEntrySList(). The latter two are different from InterlockedPopEntrySList() and InterlockedPushEntrySList() in that they accept a spinlock to hold while executing, whereas the non-Ex routines use a lock internal to ntoskrnl. I also modified ExInitializePagedLookasideList() and ExInitializeNPagedLookasideList() to initialize mutex locks within the lookaside structures. It seems that in NDIS 5.0, the lookaside allocate/free routines ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(), which require the use of the per-lookaside spinlock, whereas in NDIS 5.1, the per-lookaside spinlock is deprecated. We need to support both cases. Note that I appear to be doing something wrong with ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(): they don't appear to obtain proper pointers to their arguments, so I'm probably doing something wrong in terms of their calling convention (they're declared to be FASTCALL in Widnows, and I'm not sure what that means for gcc). It happens that in my stub lookaside implementation, they don't need to do any work anyway, so for now I've hacked them to always return NULL, which avoids corrupting the stack. I need to do this right though.
2003-12-12 22:35:13 +00:00
}
subr_ndis.c: - fix ndis_time() so that it returns a time based on the proper epoch (wacky though it may be) - implement NdisInitializeString() and NdisFreeString(), and add stub for NdisMRemoveMiniport() ntoskrnl_var.h: - add missing member to the general_lookaside struct (gl_listentry) subr_ntoskrnl.c: - Fix arguments to the interlocked push/pop routines: 'head' is an slist_header *, not an slist_entry * - Kludge up _fastcall support for the push/pop routines. The _fastcall convention is similar to _stdcall, except the first two available DWORD-sized arguments are passed in %ecx and %edx, respectively. One kludge for this __attribute__ ((regparm(3))), however this isn't entirely right, as it assumes %eax, %ecx and %edx will be used (regparm(2) assumes %eax and %edx). Another kludge is to declare the two fastcall-ed args as local register variables and explicitly assign them to %ecx and %edx, but experimentation showed that gcc would not guard %ecx and %edx against being clobbered. Thus, I came up with a 3rd kludge, which is to use some inline assembly of the form: void *arg1; void *arg2; __asm__("movl %%ecx, %%ecx" : "=c" (arg1)); __asm__("movl %%edx, %%edx" : "=d" (arg2)); This lets gcc know that we're going to reference %ecx and %edx and that it should make an effort not to let it get trampled. This wastes an instruction (movl %reg, %reg is a no-op) but insures proper behavior. It's possible there's a better way to do this though: this is the first time I've used inline assembler in this fashion. The above fixes to ntoskrnl_var.h an subr_ntoskrnl.c make lookaside lists work for the two drivers I have that use them, one of which is an NDIS 5.0 miniport and another which is 5.1.
2003-12-13 07:41:12 +00:00
__stdcall static void
ndis_init_string(dst, src)
ndis_unicode_string **dst;
char *src;
{
ndis_unicode_string *u;
u = malloc(sizeof(ndis_unicode_string), M_DEVBUF, M_NOWAIT);
if (u == NULL)
return;
u->nus_buf = NULL;
if (ndis_ascii_to_unicode(src, &u->nus_buf)) {
free(u, M_DEVBUF);
return;
}
u->nus_len = u->nus_maxlen = strlen(src) * 2;
return;
}
__stdcall static void
ndis_free_string(str)
ndis_unicode_string *str;
{
if (str == NULL)
return;
if (str->nus_buf != NULL)
free(str->nus_buf, M_DEVBUF);
free(str, M_DEVBUF);
return;
}
__stdcall static ndis_status
ndis_remove_miniport(adapter)
ndis_handle *adapter;
{
return(NDIS_STATUS_SUCCESS);
}
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void
ndis_init_ansi_string(dst, src)
ndis_ansi_string *dst;
char *src;
{
ndis_ansi_string *a;
a = dst;
if (a == NULL)
return;
if (src == NULL) {
a->nas_len = a->nas_maxlen = 0;
a->nas_buf = NULL;
} else {
a->nas_buf = src;
a->nas_len = a->nas_maxlen = strlen(src);
}
return;
}
Implement NdisInitUnicodeString().
2003-12-28 21:36:03 +00:00
__stdcall static void
ndis_init_unicode_string(dst, src)
ndis_unicode_string *dst;
uint16_t *src;
{
ndis_unicode_string *u;
int i;
u = dst;
if (u == NULL)
return;
if (src == NULL) {
u->nus_len = u->nus_maxlen = 0;
u->nus_buf = NULL;
} else {
i = 0;
while(src[i] != 0)
i++;
u->nus_buf = src;
u->nus_len = u->nus_maxlen = i * 2;
}
return;
}
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
__stdcall static void ndis_get_devprop(adapter, phydevobj,
funcdevobj, nextdevobj, resources, transresources)
ndis_handle adapter;
void *phydevobj;
void *funcdevobj;
void *nextdevobj;
cm_resource_list *resources;
cm_resource_list *transresources;
{
return;
}
Now that I finally have power back, implement a couple more NDIS API routines: NdisUnchainBufferAtBack(), NdisGetFirstBufferFromPacketSafe() and NdisGetFirstBufferFromPacket(). This should bring us a little closer to getting the Intel centrino wireless NIC to work. Note: I have not actually tested these additions since I don't have a driver that calls them, however they're pretty simple, and one of them is taken pretty much directly from the Windows ndis.h header file, so I'm fairly confident they work, but disclaimers apply.
2003-12-22 08:24:32 +00:00
__stdcall static void
ndis_firstbuf(packet, buf, firstva, firstlen, totlen)
ndis_packet *packet;
ndis_buffer **buf;
void **firstva;
uint32_t *firstlen;
uint32_t *totlen;
{
ndis_buffer *tmp;
tmp = packet->np_private.npp_head;
*buf = tmp;
if (tmp == NULL) {
*firstva = NULL;
*firstlen = *totlen = 0;
} else {
Re-do the handling of ndis_buffers. The NDIS_BUFFER structure is supposed to be opaque to the driver, however it is exposed through several macros which expect certain behavior. In my original implementation, I used the mappedsystemva member of the structure to hold a pointer to the buffer and bytecount to hold the length. It turns out you must use the startva pointer to point to the page containing the start of the buffer and set byteoffset to the offset within the page where the buffer starts. So, for a buffer with address 'baseva,' startva is baseva & ~(PAGE_SIZE -1) and byteoffset is baseva & (PAGE_SIZE -1). We have to maintain this convention everywhere that ndis_buffers are used. Fortunately, Microsoft defines some macros for initializing and manipulating NDIS_BUFFER structures in ntddk.h. I adapted some of them for use here and used them where appropriate. This fixes the discrepancy I observed between how RX'ed packet sizes were being reported in the Broadcom wireless driver and the sample ethernet drivers that I've tested. This should also help the Intel Centrino wireless driver work. Also try to properly initialize the 802.11 BSS and IBSS channels. (Sadly, the channel value is meaningless since there's no way in the existing NDIS API to get/set the channel, but this should take care of any 'invalid channel (NULL)' messages printed on the console.
2003-12-23 04:08:22 +00:00
*firstva = MDL_VA(tmp);
Now that I finally have power back, implement a couple more NDIS API routines: NdisUnchainBufferAtBack(), NdisGetFirstBufferFromPacketSafe() and NdisGetFirstBufferFromPacket(). This should bring us a little closer to getting the Intel centrino wireless NIC to work. Note: I have not actually tested these additions since I don't have a driver that calls them, however they're pretty simple, and one of them is taken pretty much directly from the Windows ndis.h header file, so I'm fairly confident they work, but disclaimers apply.
2003-12-22 08:24:32 +00:00
*firstlen = *totlen = tmp->nb_bytecount;
for (tmp = tmp->nb_next; tmp != NULL; tmp = tmp->nb_next)
*totlen += tmp->nb_bytecount;
}
return;
}
__stdcall static void
ndis_firstbuf_safe(packet, buf, firstva, firstlen, totlen, prio)
ndis_packet *packet;
ndis_buffer **buf;
void **firstva;
uint32_t *firstlen;
uint32_t *totlen;
uint32_t prio;
{
ndis_firstbuf(packet, buf, firstva, firstlen, totlen);
}
- Add stubs for Ndis*File() functions - Fix ndis_time(). - Implement NdisGetSystemUpTime(). - Implement RtlCopyUnicodeString() and RtlUnicodeStringToAnsiString(). - In ndis_getstate_80211(), use sc->ndis_link to determine connect status. Submitted by: Brian Feldman <green@freebsd.org>
2003-12-25 00:40:02 +00:00
/* can also return NDIS_STATUS_RESOURCES/NDIS_STATUS_ERROR_READING_FILE */
__stdcall static void
ndis_open_file(status, filehandle, filelength, filename, highestaddr)
ndis_status *status;
ndis_handle *filehandle;
uint32_t *filelength;
ndis_unicode_string *filename;
ndis_physaddr highestaddr;
{
char *afilename = NULL;
ndis_unicode_to_ascii(filename->nus_buf, filename->nus_len, &afilename);
printf("ndis_open_file(\"%s\", %ju)\n", afilename,
highestaddr.np_quad);
free(afilename, M_DEVBUF);
*status = NDIS_STATUS_FILE_NOT_FOUND;
return;
}
__stdcall static void
ndis_map_file(status, mappedbuffer, filehandle)
ndis_status *status;
void **mappedbuffer;
ndis_handle filehandle;
{
*status = NDIS_STATUS_ALREADY_MAPPED;
return;
}
__stdcall static void
ndis_unmap_file(filehandle)
ndis_handle filehandle;
{
return;
}
__stdcall static void
ndis_close_file(filehandle)
ndis_handle filehandle;
{
return;
}
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
__stdcall static uint8_t
ndis_cpu_cnt()
{
#ifdef SMP
return(mp_ncpus);
#else
return(1);
#endif
};
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
__stdcall static void
dummy()
{
printf ("NDIS dummy called...\n");
return;
}
image_patch_table ndis_functbl[] = {
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
{ "NdisSystemProcessorCount", (FUNC)ndis_cpu_cnt },
Now that I finally have power back, implement a couple more NDIS API routines: NdisUnchainBufferAtBack(), NdisGetFirstBufferFromPacketSafe() and NdisGetFirstBufferFromPacket(). This should bring us a little closer to getting the Intel centrino wireless NIC to work. Note: I have not actually tested these additions since I don't have a driver that calls them, however they're pretty simple, and one of them is taken pretty much directly from the Windows ndis.h header file, so I'm fairly confident they work, but disclaimers apply.
2003-12-22 08:24:32 +00:00
{ "NdisUnchainBufferAtBack", (FUNC)ndis_unchain_tailbuf, },
{ "NdisGetFirstBufferFromPacket", (FUNC)ndis_firstbuf },
{ "NdisGetFirstBufferFromPacketSafe", (FUNC)ndis_firstbuf_safe },
Implement NdisGetBufferPhysicalArraySize(), which apparently is a synonym for NDIS_BUFFER_TO_SPAN_PAGES().
2003-12-16 18:56:33 +00:00
{ "NdisGetBufferPhysicalArraySize", (FUNC)ndis_buf_physpages },
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
{ "NdisMGetDeviceProperty", (FUNC)ndis_get_devprop },
{ "NdisInitAnsiString", (FUNC)ndis_init_ansi_string },
Implement NdisInitUnicodeString().
2003-12-28 21:36:03 +00:00
{ "NdisInitUnicodeString", (FUNC)ndis_init_unicode_string },
Implement a few new NDIS routines: NdisInitAnsiString(), NdisAnsiStringToUnicodeString(), NdisWriteConfiguration(). Also add stubs for NdisMGetDeviceProperty(), NdisTerminateWrapper(), NdisOpenConfigurationKeyByName(), NdisOpenConfigurationKeyByIndex() and NdisMGetDeviceProperty().
2003-12-14 00:43:40 +00:00
{ "NdisWriteConfiguration", (FUNC)ndis_write_cfg },
{ "NdisAnsiStringToUnicodeString", (FUNC)ndis_ansi2unicode },
{ "NdisTerminateWrapper", (FUNC)ndis_termwrap },
{ "NdisOpenConfigurationKeyByName", (FUNC)ndis_open_cfgbyname },
{ "NdisOpenConfigurationKeyByIndex", (FUNC)ndis_open_cfgbyidx },
subr_ndis.c: - fix ndis_time() so that it returns a time based on the proper epoch (wacky though it may be) - implement NdisInitializeString() and NdisFreeString(), and add stub for NdisMRemoveMiniport() ntoskrnl_var.h: - add missing member to the general_lookaside struct (gl_listentry) subr_ntoskrnl.c: - Fix arguments to the interlocked push/pop routines: 'head' is an slist_header *, not an slist_entry * - Kludge up _fastcall support for the push/pop routines. The _fastcall convention is similar to _stdcall, except the first two available DWORD-sized arguments are passed in %ecx and %edx, respectively. One kludge for this __attribute__ ((regparm(3))), however this isn't entirely right, as it assumes %eax, %ecx and %edx will be used (regparm(2) assumes %eax and %edx). Another kludge is to declare the two fastcall-ed args as local register variables and explicitly assign them to %ecx and %edx, but experimentation showed that gcc would not guard %ecx and %edx against being clobbered. Thus, I came up with a 3rd kludge, which is to use some inline assembly of the form: void *arg1; void *arg2; __asm__("movl %%ecx, %%ecx" : "=c" (arg1)); __asm__("movl %%edx, %%edx" : "=d" (arg2)); This lets gcc know that we're going to reference %ecx and %edx and that it should make an effort not to let it get trampled. This wastes an instruction (movl %reg, %reg is a no-op) but insures proper behavior. It's possible there's a better way to do this though: this is the first time I've used inline assembler in this fashion. The above fixes to ntoskrnl_var.h an subr_ntoskrnl.c make lookaside lists work for the two drivers I have that use them, one of which is an NDIS 5.0 miniport and another which is 5.1.
2003-12-13 07:41:12 +00:00
{ "NdisMRemoveMiniport", (FUNC)ndis_remove_miniport },
{ "NdisInitializeString", (FUNC)ndis_init_string },
{ "NdisFreeString", (FUNC)ndis_free_string },
Implement some more NDIS and ntoskrnl API calls: subr_ndis.c: NdisGetCurrentSystemTime() which, according to the Microsoft documentation returns "the number of 100 nanosecond intervals since January 1, 1601." I have no idea what's so special about that epoch or why they chose 100 nanosecond ticks. I don't know the proper offset to convert nanotime() from the UNIX epoch to January 1, 1601, so for now I'm just doing the unit convertion to 100s of nanoseconds. subr_ntoskrnl.c: memcpy(), memset(), ExInterlockedPopEntrySList(), ExInterlockedPushEntrySList(). The latter two are different from InterlockedPopEntrySList() and InterlockedPushEntrySList() in that they accept a spinlock to hold while executing, whereas the non-Ex routines use a lock internal to ntoskrnl. I also modified ExInitializePagedLookasideList() and ExInitializeNPagedLookasideList() to initialize mutex locks within the lookaside structures. It seems that in NDIS 5.0, the lookaside allocate/free routines ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(), which require the use of the per-lookaside spinlock, whereas in NDIS 5.1, the per-lookaside spinlock is deprecated. We need to support both cases. Note that I appear to be doing something wrong with ExInterlockedPopEntrySList() and ExInterlockedPushEntrySList(): they don't appear to obtain proper pointers to their arguments, so I'm probably doing something wrong in terms of their calling convention (they're declared to be FASTCALL in Widnows, and I'm not sure what that means for gcc). It happens that in my stub lookaside implementation, they don't need to do any work anyway, so for now I've hacked them to always return NULL, which avoids corrupting the stack. I need to do this right though.
2003-12-12 22:35:13 +00:00
{ "NdisGetCurrentSystemTime", (FUNC)ndis_time },
- Add stubs for Ndis*File() functions - Fix ndis_time(). - Implement NdisGetSystemUpTime(). - Implement RtlCopyUnicodeString() and RtlUnicodeStringToAnsiString(). - In ndis_getstate_80211(), use sc->ndis_link to determine connect status. Submitted by: Brian Feldman <green@freebsd.org>
2003-12-25 00:40:02 +00:00
{ "NdisGetSystemUpTime", (FUNC)ndis_uptime },
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
{ "NdisMSynchronizeWithInterrupt", (FUNC)ndis_sync_with_intr },
{ "NdisMAllocateSharedMemoryAsync", (FUNC)ndis_alloc_sharedmem_async },
{ "NdisInterlockedInsertHeadList", (FUNC)ndis_insert_head },
{ "NdisInterlockedInsertTailList", (FUNC)ndis_insert_tail },
{ "NdisInterlockedRemoveHeadList", (FUNC)ndis_remove_head },
{ "NdisInitializeWrapper", (FUNC)ndis_initwrap },
{ "NdisMRegisterMiniport", (FUNC)ndis_register_miniport },
{ "NdisAllocateMemoryWithTag", (FUNC)ndis_malloc_withtag },
{ "NdisAllocateMemory", (FUNC)ndis_malloc },
{ "NdisMSetAttributesEx", (FUNC)ndis_setattr_ex },
{ "NdisCloseConfiguration", (FUNC)ndis_close_cfg },
{ "NdisReadConfiguration", (FUNC)ndis_read_cfg },
{ "NdisOpenConfiguration", (FUNC)ndis_open_cfg },
{ "NdisReleaseSpinLock", (FUNC)ndis_unlock },
{ "NdisDprAcquireSpinLock", (FUNC)ndis_lock },
{ "NdisDprReleaseSpinLock", (FUNC)ndis_unlock },
{ "NdisAcquireSpinLock", (FUNC)ndis_lock },
{ "NdisAllocateSpinLock", (FUNC)ndis_create_lock },
{ "NdisFreeSpinLock", (FUNC)ndis_destroy_lock },
{ "NdisFreeMemory", (FUNC)ndis_free },
{ "NdisReadPciSlotInformation", (FUNC)ndis_read_pci },
{ "NdisWritePciSlotInformation",(FUNC)ndis_write_pci },
{ "NdisWriteErrorLogEntry", (FUNC)ndis_syslog },
{ "NdisMStartBufferPhysicalMapping", (FUNC)ndis_vtophys_load },
{ "NdisMCompleteBufferPhysicalMapping", (FUNC)ndis_vtophys_unload },
{ "NdisMInitializeTimer", (FUNC)ndis_create_timer },
{ "NdisSetTimer", (FUNC)ndis_set_timer },
{ "NdisMCancelTimer", (FUNC)ndis_cancel_timer },
{ "NdisMSetPeriodicTimer", (FUNC)ndis_set_periodic_timer },
{ "NdisMQueryAdapterResources", (FUNC)ndis_query_resources },
{ "NdisMRegisterIoPortRange", (FUNC)ndis_register_ioport },
{ "NdisMDeregisterIoPortRange", (FUNC)ndis_deregister_ioport },
{ "NdisReadNetworkAddress", (FUNC)ndis_read_netaddr },
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
{ "NdisQueryMapRegisterCount", (FUNC)ndis_mapreg_cnt },
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
{ "NdisMAllocateMapRegisters", (FUNC)ndis_alloc_mapreg },
Back out the last batch of changes until I have a chance to properly evaluate them. Whatever they're meant to do, they're doing it wrong. Also: - Clean up last bits of NULL fallout in subr_pe - Don't let ndis_ifmedia_sts() do anything if the IFF_UP flag isn't set - Implement NdisSystemProcessorCount() and NdisQueryMapRegisterCount().
2003-12-26 03:31:34 +00:00
{ "NdisMFreeMapRegisters", (FUNC)ndis_free_mapreg },
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
{ "NdisMAllocateSharedMemory", (FUNC)ndis_alloc_sharedmem },
{ "NdisMMapIoSpace", (FUNC)ndis_map_iospace },
{ "NdisMUnmapIoSpace", (FUNC)ndis_unmap_iospace },
{ "NdisGetCacheFillSize", (FUNC)ndis_cachefill },
{ "NdisMGetDmaAlignment", (FUNC)ndis_dma_align },
{ "NdisMInitializeScatterGatherDma", (FUNC)ndis_init_sc_dma },
{ "NdisAllocatePacketPool", (FUNC)ndis_alloc_packetpool },
{ "NdisAllocatePacketPoolEx", (FUNC)ndis_ex_alloc_packetpool },
{ "NdisAllocatePacket", (FUNC)ndis_alloc_packet },
{ "NdisFreePacket", (FUNC)ndis_release_packet },
{ "NdisFreePacketPool", (FUNC)ndis_free_packetpool },
{ "NdisAllocateBufferPool", (FUNC)ndis_alloc_bufpool },
{ "NdisAllocateBuffer", (FUNC)ndis_alloc_buf },
{ "NdisQueryBuffer", (FUNC)ndis_query_buf },
{ "NdisQueryBufferSafe", (FUNC)ndis_query_buf_safe },
{ "NdisFreeBuffer", (FUNC)ndis_release_buf },
{ "NdisFreeBufferPool", (FUNC)ndis_free_bufpool },
{ "NdisInterlockedIncrement", (FUNC)ndis_interlock_inc },
{ "NdisInterlockedDecrement", (FUNC)ndis_interlock_dec },
{ "NdisInitializeEvent", (FUNC)ndis_init_event },
{ "NdisSetEvent", (FUNC)ndis_set_event },
{ "NdisResetEvent", (FUNC)ndis_reset_event },
{ "NdisWaitEvent", (FUNC)ndis_wait_event },
{ "NdisUnicodeStringToAnsiString", (FUNC)ndis_unicode2ansi },
{ "NdisMPciAssignResources", (FUNC)ndis_assign_pcirsrc },
{ "NdisMFreeSharedMemory", (FUNC)ndis_free_sharedmem },
{ "NdisMRegisterInterrupt", (FUNC)ndis_register_intr },
{ "NdisMDeregisterInterrupt", (FUNC)ndis_deregister_intr },
{ "NdisMRegisterAdapterShutdownHandler", (FUNC)ndis_register_shutdown },
{ "NdisMDeregisterAdapterShutdownHandler", (FUNC)ndis_deregister_shutdown },
{ "NDIS_BUFFER_TO_SPAN_PAGES", (FUNC)ndis_numpages },
{ "NdisQueryBufferOffset", (FUNC)ndis_query_bufoffset },
{ "NdisAdjustBufferLength", (FUNC)ndis_adjust_buflen },
{ "NdisPacketPoolUsage", (FUNC)ndis_packetpool_use },
{ "NdisMSleep", (FUNC)ndis_sleep },
{ "NdisUnchainBufferAtFront", (FUNC)ndis_unchain_headbuf },
{ "NdisReadPcmciaAttributeMemory", (FUNC)ndis_read_pccard_amem },
{ "NdisWritePcmciaAttributeMemory", (FUNC)ndis_write_pccard_amem },
- Add stubs for Ndis*File() functions - Fix ndis_time(). - Implement NdisGetSystemUpTime(). - Implement RtlCopyUnicodeString() and RtlUnicodeStringToAnsiString(). - In ndis_getstate_80211(), use sc->ndis_link to determine connect status. Submitted by: Brian Feldman <green@freebsd.org>
2003-12-25 00:40:02 +00:00
{ "NdisOpenFile", (FUNC)ndis_open_file },
{ "NdisMapFile", (FUNC)ndis_map_file },
{ "NdisUnmapFile", (FUNC)ndis_unmap_file },
{ "NdisCloseFile", (FUNC)ndis_close_file },
Commit the first cut of Project Evil, also known as the NDISulator. Yes, it's what you think it is. Yes, you should run away now. This is a special compatibility module for allowing Windows NDIS miniport network drivers to be used with FreeBSD/x86. This provides _binary_ NDIS compatibility (not source): you can run NDIS driver code, but you can't build it. There are three main parts: sys/compat/ndis: the NDIS compat API, which provides binary compatibility functions for many routines in NDIS.SYS, HAL.dll and ntoskrnl.exe in Windows (these are the three modules that most NDIS miniport drivers use). The compat module also contains a small PE relocator/dynalinker which relocates the Windows .SYS image and then patches in our native routines. sys/dev/if_ndis: the if_ndis driver wrapper. This module makes use of the ndis compat API and can be compiled with a specially prepared binary image file (ndis_driver_data.h) containing the Windows .SYS image and registry key information parsed out of the accompanying .INF file. Once if_ndis.ko is built, it can be loaded and unloaded just like a native FreeBSD kenrel module. usr.sbin/ndiscvt: a special utility that converts foo.sys and foo.inf into an ndis_driver_data.h file that can be compiled into if_ndis.o. Contains an .inf file parser graciously provided by Matt Dodd (and mercilessly hacked upon by me) that strips out device ID info and registry key info from a .INF file and packages it up with a binary image array. The ndiscvt(8) utility also does some manipulation of the segments within the .sys file to make life easier for the kernel loader. (Doing the manipulation here saves the kernel code from having to move things around later, which would waste memory.) ndiscvt is only built for the i386 arch. Only files.i386 has been updated, and none of this is turned on in GENERIC. It should probably work on pc98. I have no idea about amd64 or ia64 at this point. This is still a work in progress. I estimate it's about %85 done, but I want it under CVS control so I can track subsequent changes. It has been tested with exactly three drivers: the LinkSys LNE100TX v4 driver (Lne100v4.sys), the sample Intel 82559 driver from the Windows DDK (e100bex.sys) and the Broadcom BCM43xx wireless driver (bcmwl5.sys). It still needs to have a net80211 stuff added to it. To use it, you would do something like this: # cd /sys/modules/ndis # make; make load # cd /sys/modules/if_ndis # ndiscvt -i /path/to/foo.inf -s /path/to/foo.sys -o ndis_driver_data.h # make; make load # sysctl -a | grep ndis All registry keys are mapped to sysctl nodes. Sometimes drivers refer to registry keys that aren't mentioned in foo.inf. If this happens, the NDIS API module creates sysctl nodes for these keys on the fly so you can tweak them. An example usage of the Broadcom wireless driver would be: # sysctl hw.ndis0.EnableAutoConnect=1 # sysctl hw.ndis0.SSID="MY_SSID" # sysctl hw.ndis0.NetworkType=0 (0 for bss, 1 for adhoc) # ifconfig ndis0 <my ipaddr> netmask 0xffffff00 up Things to be done: - get rid of debug messages - add in ndis80211 support - defer transmissions until after a status update with NDIS_STATUS_CONNECTED occurs - Create smarter lookaside list support - Split off if_ndis_pci.c and if_ndis_pccard.c attachments - Make sure PCMCIA support works - Fix ndiscvt to properly parse PCMCIA device IDs from INF files - write ndisapi.9 man page
2003-12-11 22:34:37 +00:00
/*
* This last entry is a catch-all for any function we haven't
* implemented yet. The PE import list patching routine will
* use it for any function that doesn't have an explicit match
* in this table.
*/
{ NULL, (FUNC)dummy },
/* End of list. */
{ NULL, NULL },
};
Reference in New Issue Copy Permalink