d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
467b3975a3
freebsd-skq/sys/kern/subr_bus_dma.c

560 lines
14 KiB
C
Raw Normal View History

Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
/*-
* Copyright (c) 2012 EMC Corp.
* All rights reserved.
*
* Copyright (c) 1997, 1998 Justin T. Gibbs.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_bus.h"
#include <sys/param.h>
#include <sys/conf.h>
#include <sys/systm.h>
#include <sys/bio.h>
#include <sys/bus.h>
#include <sys/callout.h>
#include <sys/mbuf.h>
#include <sys/memdesc.h>
#include <sys/proc.h>
#include <sys/uio.h>
#include <vm/vm.h>
#include <vm/vm_page.h>
#include <vm/vm_map.h>
#include <vm/pmap.h>
#include <cam/cam.h>
#include <cam/cam_ccb.h>
#include <machine/bus.h>
/*
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
* Load up data starting at offset within a region specified by a
* list of virtual address ranges until either length or the region
* are exhausted.
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
*/
static int
_bus_dmamap_load_vlist(bus_dma_tag_t dmat, bus_dmamap_t map,
bus_dma_segment_t *list, int sglist_cnt, struct pmap *pmap, int *nsegs,
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
int flags, size_t offset, size_t length)
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
{
int error;
error = 0;
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
for (; sglist_cnt > 0 && length != 0; sglist_cnt--, list++) {
char *addr;
size_t ds_len;
KASSERT((offset < list->ds_len),
("Invalid mid-segment offset"));
addr = (char *)(uintptr_t)list->ds_addr + offset;
ds_len = list->ds_len - offset;
offset = 0;
if (ds_len > length)
ds_len = length;
length -= ds_len;
KASSERT((ds_len != 0), ("Segment length is zero"));
error = _bus_dmamap_load_buffer(dmat, map, addr, ds_len, pmap,
Some compilers issue a warning when wider integer is casted to narrow pointer. Supposedly shut down the warning by casting through uintptr_t. Reported by: ian
2013-04-16 07:11:52 +00:00
flags, NULL, nsegs);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
if (error)
break;
}
return (error);
}
/*
* Load a list of physical addresses.
*/
static int
_bus_dmamap_load_plist(bus_dma_tag_t dmat, bus_dmamap_t map,
bus_dma_segment_t *list, int sglist_cnt, int *nsegs, int flags)
{
int error;
error = 0;
for (; sglist_cnt > 0; sglist_cnt--, list++) {
error = _bus_dmamap_load_phys(dmat, map,
(vm_paddr_t)list->ds_addr, list->ds_len, flags, NULL,
nsegs);
if (error)
break;
}
return (error);
}
/*
* Load an mbuf chain.
*/
static int
_bus_dmamap_load_mbuf_sg(bus_dma_tag_t dmat, bus_dmamap_t map,
struct mbuf *m0, bus_dma_segment_t *segs, int *nsegs, int flags)
{
struct mbuf *m;
int error;
error = 0;
for (m = m0; m != NULL && error == 0; m = m->m_next) {
if (m->m_len > 0) {
error = _bus_dmamap_load_buffer(dmat, map, m->m_data,
m->m_len, kernel_pmap, flags | BUS_DMA_LOAD_MBUF,
segs, nsegs);
}
}
CTR5(KTR_BUSDMA, "%s: tag %p tag flags 0x%x error %d nsegs %d",
__func__, dmat, flags, error, *nsegs);
return (error);
}
/*
* Load from block io.
*/
static int
_bus_dmamap_load_bio(bus_dma_tag_t dmat, bus_dmamap_t map, struct bio *bio,
int *nsegs, int flags)
{
Implement the concept of the unmapped VMIO buffers, i.e. buffers which do not map the b_pages pages into buffer_map KVA. The use of the unmapped buffers eliminate the need to perform TLB shootdown for mapping on the buffer creation and reuse, greatly reducing the amount of IPIs for shootdown on big-SMP machines and eliminating up to 25-30% of the system time on i/o intensive workloads. The unmapped buffer should be explicitely requested by the GB_UNMAPPED flag by the consumer. For unmapped buffer, no KVA reservation is performed at all. The consumer might request unmapped buffer which does have a KVA reserve, to manually map it without recursing into buffer cache and blocking, with the GB_KVAALLOC flag. When the mapped buffer is requested and unmapped buffer already exists, the cache performs an upgrade, possibly reusing the KVA reservation. Unmapped buffer is translated into unmapped bio in g_vfs_strategy(). Unmapped bio carry a pointer to the vm_page_t array, offset and length instead of the data pointer. The provider which processes the bio should explicitely specify a readiness to accept unmapped bio, otherwise g_down geom thread performs the transient upgrade of the bio request by mapping the pages into the new bio_transient_map KVA submap. The bio_transient_map submap claims up to 10% of the buffer map, and the total buffer_map + bio_transient_map KVA usage stays the same. Still, it could be manually tuned by kern.bio_transient_maxcnt tunable, in the units of the transient mappings. Eventually, the bio_transient_map could be removed after all geom classes and drivers can accept unmapped i/o requests. Unmapped support can be turned off by the vfs.unmapped_buf_allowed tunable, disabling which makes the buffer (or cluster) creation requests to ignore GB_UNMAPPED and GB_KVAALLOC flags. Unmapped buffers are only enabled by default on the architectures where pmap_copy_page() was implemented and tested. In the rework, filesystem metadata is not the subject to maxbufspace limit anymore. Since the metadata buffers are always mapped, the buffers still have to fit into the buffer map, which provides a reasonable (but practically unreachable) upper bound on it. The non-metadata buffer allocations, both mapped and unmapped, is accounted against maxbufspace, as before. Effectively, this means that the maxbufspace is forced on mapped and unmapped buffers separately. The pre-patch bufspace limiting code did not worked, because buffer_map fragmentation does not allow the limit to be reached. By Jeff Roberson request, the getnewbuf() function was split into smaller single-purpose functions. Sponsored by: The FreeBSD Foundation Discussed with: jeff (previous version) Tested by: pho, scottl (previous version), jhb, bf MFC after: 2 weeks
2013-03-19 14:13:12 +00:00
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
if ((bio->bio_flags & BIO_VLIST) != 0) {
bus_dma_segment_t *segs = (bus_dma_segment_t *)bio->bio_data;
return (_bus_dmamap_load_vlist(dmat, map, segs, bio->bio_ma_n,
kernel_pmap, nsegs, flags, bio->bio_ma_offset,
bio->bio_bcount));
Implement the concept of the unmapped VMIO buffers, i.e. buffers which do not map the b_pages pages into buffer_map KVA. The use of the unmapped buffers eliminate the need to perform TLB shootdown for mapping on the buffer creation and reuse, greatly reducing the amount of IPIs for shootdown on big-SMP machines and eliminating up to 25-30% of the system time on i/o intensive workloads. The unmapped buffer should be explicitely requested by the GB_UNMAPPED flag by the consumer. For unmapped buffer, no KVA reservation is performed at all. The consumer might request unmapped buffer which does have a KVA reserve, to manually map it without recursing into buffer cache and blocking, with the GB_KVAALLOC flag. When the mapped buffer is requested and unmapped buffer already exists, the cache performs an upgrade, possibly reusing the KVA reservation. Unmapped buffer is translated into unmapped bio in g_vfs_strategy(). Unmapped bio carry a pointer to the vm_page_t array, offset and length instead of the data pointer. The provider which processes the bio should explicitely specify a readiness to accept unmapped bio, otherwise g_down geom thread performs the transient upgrade of the bio request by mapping the pages into the new bio_transient_map KVA submap. The bio_transient_map submap claims up to 10% of the buffer map, and the total buffer_map + bio_transient_map KVA usage stays the same. Still, it could be manually tuned by kern.bio_transient_maxcnt tunable, in the units of the transient mappings. Eventually, the bio_transient_map could be removed after all geom classes and drivers can accept unmapped i/o requests. Unmapped support can be turned off by the vfs.unmapped_buf_allowed tunable, disabling which makes the buffer (or cluster) creation requests to ignore GB_UNMAPPED and GB_KVAALLOC flags. Unmapped buffers are only enabled by default on the architectures where pmap_copy_page() was implemented and tested. In the rework, filesystem metadata is not the subject to maxbufspace limit anymore. Since the metadata buffers are always mapped, the buffers still have to fit into the buffer map, which provides a reasonable (but practically unreachable) upper bound on it. The non-metadata buffer allocations, both mapped and unmapped, is accounted against maxbufspace, as before. Effectively, this means that the maxbufspace is forced on mapped and unmapped buffers separately. The pre-patch bufspace limiting code did not worked, because buffer_map fragmentation does not allow the limit to be reached. By Jeff Roberson request, the getnewbuf() function was split into smaller single-purpose functions. Sponsored by: The FreeBSD Foundation Discussed with: jeff (previous version) Tested by: pho, scottl (previous version), jhb, bf MFC after: 2 weeks
2013-03-19 14:13:12 +00:00
}
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
if ((bio->bio_flags & BIO_UNMAPPED) != 0)
Fix a bug introduced in r291716: "The problem with the approach taken both in _bus_dmamap_load_pages and bus_dmamap_load_ma_triv is that they split the request buffer into arbitrary chunks based on page boundaries, creating segments that no longer have a size that's a multiple of the sector size. This breaks drivers like blkfront (and probably other stuff)." [1] This was most easily triggered by running `fsck /` on a system running in Xen (e.g. Amazon EC2) but also showed up via growfs(8) and probably many other userland tools which access the disk directly. Patch by: royger [1] "Thinks this should be fine" by: ken
2016-01-11 20:38:39 +00:00
return (_bus_dmamap_load_ma(dmat, map, bio->bio_ma,
bio->bio_bcount, bio->bio_ma_offset, flags, NULL, nsegs));
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
return (_bus_dmamap_load_buffer(dmat, map, bio->bio_data,
bio->bio_bcount, kernel_pmap, flags, NULL, nsegs));
Add bus_dmamap_load_ma() function to load map with the array of vm_pages. Provide trivial implementation which forwards the load to _bus_dmamap_load_phys() page by page. Right now all architectures use bus_dmamap_load_ma_triv(). Tested by: pho (as part of the functional patch) Sponsored by: The FreeBSD Foundation MFC after: 1 month
2013-10-27 21:39:16 +00:00
}
int
bus_dmamap_load_ma_triv(bus_dma_tag_t dmat, bus_dmamap_t map,
struct vm_page **ma, bus_size_t tlen, int ma_offs, int flags,
bus_dma_segment_t *segs, int *segp)
{
vm_paddr_t paddr;
bus_size_t len;
int error, i;
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
Implement the concept of the unmapped VMIO buffers, i.e. buffers which do not map the b_pages pages into buffer_map KVA. The use of the unmapped buffers eliminate the need to perform TLB shootdown for mapping on the buffer creation and reuse, greatly reducing the amount of IPIs for shootdown on big-SMP machines and eliminating up to 25-30% of the system time on i/o intensive workloads. The unmapped buffer should be explicitely requested by the GB_UNMAPPED flag by the consumer. For unmapped buffer, no KVA reservation is performed at all. The consumer might request unmapped buffer which does have a KVA reserve, to manually map it without recursing into buffer cache and blocking, with the GB_KVAALLOC flag. When the mapped buffer is requested and unmapped buffer already exists, the cache performs an upgrade, possibly reusing the KVA reservation. Unmapped buffer is translated into unmapped bio in g_vfs_strategy(). Unmapped bio carry a pointer to the vm_page_t array, offset and length instead of the data pointer. The provider which processes the bio should explicitely specify a readiness to accept unmapped bio, otherwise g_down geom thread performs the transient upgrade of the bio request by mapping the pages into the new bio_transient_map KVA submap. The bio_transient_map submap claims up to 10% of the buffer map, and the total buffer_map + bio_transient_map KVA usage stays the same. Still, it could be manually tuned by kern.bio_transient_maxcnt tunable, in the units of the transient mappings. Eventually, the bio_transient_map could be removed after all geom classes and drivers can accept unmapped i/o requests. Unmapped support can be turned off by the vfs.unmapped_buf_allowed tunable, disabling which makes the buffer (or cluster) creation requests to ignore GB_UNMAPPED and GB_KVAALLOC flags. Unmapped buffers are only enabled by default on the architectures where pmap_copy_page() was implemented and tested. In the rework, filesystem metadata is not the subject to maxbufspace limit anymore. Since the metadata buffers are always mapped, the buffers still have to fit into the buffer map, which provides a reasonable (but practically unreachable) upper bound on it. The non-metadata buffer allocations, both mapped and unmapped, is accounted against maxbufspace, as before. Effectively, this means that the maxbufspace is forced on mapped and unmapped buffers separately. The pre-patch bufspace limiting code did not worked, because buffer_map fragmentation does not allow the limit to be reached. By Jeff Roberson request, the getnewbuf() function was split into smaller single-purpose functions. Sponsored by: The FreeBSD Foundation Discussed with: jeff (previous version) Tested by: pho, scottl (previous version), jhb, bf MFC after: 2 weeks
2013-03-19 14:13:12 +00:00
error = 0;
for (i = 0; tlen > 0; i++, tlen -= len) {
len = min(PAGE_SIZE - ma_offs, tlen);
Add bus_dmamap_load_ma() function to load map with the array of vm_pages. Provide trivial implementation which forwards the load to _bus_dmamap_load_phys() page by page. Right now all architectures use bus_dmamap_load_ma_triv(). Tested by: pho (as part of the functional patch) Sponsored by: The FreeBSD Foundation MFC after: 1 month
2013-10-27 21:39:16 +00:00
paddr = VM_PAGE_TO_PHYS(ma[i]) + ma_offs;
Implement the concept of the unmapped VMIO buffers, i.e. buffers which do not map the b_pages pages into buffer_map KVA. The use of the unmapped buffers eliminate the need to perform TLB shootdown for mapping on the buffer creation and reuse, greatly reducing the amount of IPIs for shootdown on big-SMP machines and eliminating up to 25-30% of the system time on i/o intensive workloads. The unmapped buffer should be explicitely requested by the GB_UNMAPPED flag by the consumer. For unmapped buffer, no KVA reservation is performed at all. The consumer might request unmapped buffer which does have a KVA reserve, to manually map it without recursing into buffer cache and blocking, with the GB_KVAALLOC flag. When the mapped buffer is requested and unmapped buffer already exists, the cache performs an upgrade, possibly reusing the KVA reservation. Unmapped buffer is translated into unmapped bio in g_vfs_strategy(). Unmapped bio carry a pointer to the vm_page_t array, offset and length instead of the data pointer. The provider which processes the bio should explicitely specify a readiness to accept unmapped bio, otherwise g_down geom thread performs the transient upgrade of the bio request by mapping the pages into the new bio_transient_map KVA submap. The bio_transient_map submap claims up to 10% of the buffer map, and the total buffer_map + bio_transient_map KVA usage stays the same. Still, it could be manually tuned by kern.bio_transient_maxcnt tunable, in the units of the transient mappings. Eventually, the bio_transient_map could be removed after all geom classes and drivers can accept unmapped i/o requests. Unmapped support can be turned off by the vfs.unmapped_buf_allowed tunable, disabling which makes the buffer (or cluster) creation requests to ignore GB_UNMAPPED and GB_KVAALLOC flags. Unmapped buffers are only enabled by default on the architectures where pmap_copy_page() was implemented and tested. In the rework, filesystem metadata is not the subject to maxbufspace limit anymore. Since the metadata buffers are always mapped, the buffers still have to fit into the buffer map, which provides a reasonable (but practically unreachable) upper bound on it. The non-metadata buffer allocations, both mapped and unmapped, is accounted against maxbufspace, as before. Effectively, this means that the maxbufspace is forced on mapped and unmapped buffers separately. The pre-patch bufspace limiting code did not worked, because buffer_map fragmentation does not allow the limit to be reached. By Jeff Roberson request, the getnewbuf() function was split into smaller single-purpose functions. Sponsored by: The FreeBSD Foundation Discussed with: jeff (previous version) Tested by: pho, scottl (previous version), jhb, bf MFC after: 2 weeks
2013-03-19 14:13:12 +00:00
error = _bus_dmamap_load_phys(dmat, map, paddr, len,
Add bus_dmamap_load_ma() function to load map with the array of vm_pages. Provide trivial implementation which forwards the load to _bus_dmamap_load_phys() page by page. Right now all architectures use bus_dmamap_load_ma_triv(). Tested by: pho (as part of the functional patch) Sponsored by: The FreeBSD Foundation MFC after: 1 month
2013-10-27 21:39:16 +00:00
flags, segs, segp);
Implement the concept of the unmapped VMIO buffers, i.e. buffers which do not map the b_pages pages into buffer_map KVA. The use of the unmapped buffers eliminate the need to perform TLB shootdown for mapping on the buffer creation and reuse, greatly reducing the amount of IPIs for shootdown on big-SMP machines and eliminating up to 25-30% of the system time on i/o intensive workloads. The unmapped buffer should be explicitely requested by the GB_UNMAPPED flag by the consumer. For unmapped buffer, no KVA reservation is performed at all. The consumer might request unmapped buffer which does have a KVA reserve, to manually map it without recursing into buffer cache and blocking, with the GB_KVAALLOC flag. When the mapped buffer is requested and unmapped buffer already exists, the cache performs an upgrade, possibly reusing the KVA reservation. Unmapped buffer is translated into unmapped bio in g_vfs_strategy(). Unmapped bio carry a pointer to the vm_page_t array, offset and length instead of the data pointer. The provider which processes the bio should explicitely specify a readiness to accept unmapped bio, otherwise g_down geom thread performs the transient upgrade of the bio request by mapping the pages into the new bio_transient_map KVA submap. The bio_transient_map submap claims up to 10% of the buffer map, and the total buffer_map + bio_transient_map KVA usage stays the same. Still, it could be manually tuned by kern.bio_transient_maxcnt tunable, in the units of the transient mappings. Eventually, the bio_transient_map could be removed after all geom classes and drivers can accept unmapped i/o requests. Unmapped support can be turned off by the vfs.unmapped_buf_allowed tunable, disabling which makes the buffer (or cluster) creation requests to ignore GB_UNMAPPED and GB_KVAALLOC flags. Unmapped buffers are only enabled by default on the architectures where pmap_copy_page() was implemented and tested. In the rework, filesystem metadata is not the subject to maxbufspace limit anymore. Since the metadata buffers are always mapped, the buffers still have to fit into the buffer map, which provides a reasonable (but practically unreachable) upper bound on it. The non-metadata buffer allocations, both mapped and unmapped, is accounted against maxbufspace, as before. Effectively, this means that the maxbufspace is forced on mapped and unmapped buffers separately. The pre-patch bufspace limiting code did not worked, because buffer_map fragmentation does not allow the limit to be reached. By Jeff Roberson request, the getnewbuf() function was split into smaller single-purpose functions. Sponsored by: The FreeBSD Foundation Discussed with: jeff (previous version) Tested by: pho, scottl (previous version), jhb, bf MFC after: 2 weeks
2013-03-19 14:13:12 +00:00
if (error != 0)
break;
ma_offs = 0;
}
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
return (error);
}
/*
* Load a cam control block.
*/
static int
_bus_dmamap_load_ccb(bus_dma_tag_t dmat, bus_dmamap_t map, union ccb *ccb,
int *nsegs, int flags)
{
struct ccb_hdr *ccb_h;
void *data_ptr;
int error;
uint32_t dxfer_len;
uint16_t sglist_cnt;
error = 0;
ccb_h = &ccb->ccb_h;
switch (ccb_h->func_code) {
Add support for XPT_CONT_TARGET_IO CCBs in _bus_dmamap_load_ccb(). Declare CCB types in their respective switch blocks. Sponsored by: Spectra Logic
2013-04-02 16:49:49 +00:00
case XPT_SCSI_IO: {
struct ccb_scsiio *csio;
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
csio = &ccb->csio;
data_ptr = csio->data_ptr;
dxfer_len = csio->dxfer_len;
sglist_cnt = csio->sglist_cnt;
break;
Add support for XPT_CONT_TARGET_IO CCBs in _bus_dmamap_load_ccb(). Declare CCB types in their respective switch blocks. Sponsored by: Spectra Logic
2013-04-02 16:49:49 +00:00
}
case XPT_CONT_TARGET_IO: {
struct ccb_scsiio *ctio;
ctio = &ccb->ctio;
data_ptr = ctio->data_ptr;
dxfer_len = ctio->dxfer_len;
sglist_cnt = ctio->sglist_cnt;
break;
}
case XPT_ATA_IO: {
struct ccb_ataio *ataio;
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
ataio = &ccb->ataio;
data_ptr = ataio->data_ptr;
dxfer_len = ataio->dxfer_len;
sglist_cnt = 0;
break;
Add support for XPT_CONT_TARGET_IO CCBs in _bus_dmamap_load_ccb(). Declare CCB types in their respective switch blocks. Sponsored by: Spectra Logic
2013-04-02 16:49:49 +00:00
}
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
default:
panic("_bus_dmamap_load_ccb: Unsupported func code %d",
ccb_h->func_code);
}
switch ((ccb_h->flags & CAM_DATA_MASK)) {
case CAM_DATA_VADDR:
error = _bus_dmamap_load_buffer(dmat, map, data_ptr, dxfer_len,
kernel_pmap, flags, NULL, nsegs);
break;
case CAM_DATA_PADDR:
error = _bus_dmamap_load_phys(dmat, map,
(vm_paddr_t)(uintptr_t)data_ptr, dxfer_len, flags, NULL,
nsegs);
break;
case CAM_DATA_SG:
error = _bus_dmamap_load_vlist(dmat, map,
(bus_dma_segment_t *)data_ptr, sglist_cnt, kernel_pmap,
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
nsegs, flags, 0, dxfer_len);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
break;
case CAM_DATA_SG_PADDR:
error = _bus_dmamap_load_plist(dmat, map,
(bus_dma_segment_t *)data_ptr, sglist_cnt, nsegs, flags);
break;
case CAM_DATA_BIO:
error = _bus_dmamap_load_bio(dmat, map, (struct bio *)data_ptr,
nsegs, flags);
break;
default:
panic("_bus_dmamap_load_ccb: flags 0x%X unimplemented",
ccb_h->flags);
}
return (error);
}
/*
* Load a uio.
*/
static int
_bus_dmamap_load_uio(bus_dma_tag_t dmat, bus_dmamap_t map, struct uio *uio,
int *nsegs, int flags)
{
bus_size_t resid;
bus_size_t minlen;
struct iovec *iov;
pmap_t pmap;
caddr_t addr;
int error, i;
if (uio->uio_segflg == UIO_USERSPACE) {
KASSERT(uio->uio_td != NULL,
("bus_dmamap_load_uio: USERSPACE but no proc"));
pmap = vmspace_pmap(uio->uio_td->td_proc->p_vmspace);
} else
pmap = kernel_pmap;
resid = uio->uio_resid;
iov = uio->uio_iov;
error = 0;
for (i = 0; i < uio->uio_iovcnt && resid != 0 && !error; i++) {
/*
* Now at the first iovec to load. Load each iovec
* until we have exhausted the residual count.
*/
addr = (caddr_t) iov[i].iov_base;
minlen = resid < iov[i].iov_len ? resid : iov[i].iov_len;
if (minlen > 0) {
error = _bus_dmamap_load_buffer(dmat, map, addr,
minlen, pmap, flags, NULL, nsegs);
resid -= minlen;
}
}
return (error);
}
/*
* Map the buffer buf into bus space using the dmamap map.
*/
int
bus_dmamap_load(bus_dma_tag_t dmat, bus_dmamap_t map, void *buf,
bus_size_t buflen, bus_dmamap_callback_t *callback,
void *callback_arg, int flags)
{
bus_dma_segment_t *segs;
struct memdesc mem;
int error;
int nsegs;
if ((flags & BUS_DMA_NOWAIT) == 0) {
mem = memdesc_vaddr(buf, buflen);
_bus_dmamap_waitok(dmat, map, &mem, callback, callback_arg);
}
nsegs = -1;
error = _bus_dmamap_load_buffer(dmat, map, buf, buflen, kernel_pmap,
flags, NULL, &nsegs);
nsegs++;
CTR5(KTR_BUSDMA, "%s: tag %p tag flags 0x%x error %d nsegs %d",
Do not add 1 to nsegs before passing to CTR5(), since nsegs has already been incremented before these calls. Sponsored by: Intel
2013-03-29 15:54:12 +00:00
__func__, dmat, flags, error, nsegs);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
if (error == EINPROGRESS)
return (error);
segs = _bus_dmamap_complete(dmat, map, NULL, nsegs, error);
if (error)
(*callback)(callback_arg, segs, 0, error);
else
(*callback)(callback_arg, segs, nsegs, 0);
/*
* Return ENOMEM to the caller so that it can pass it up the stack.
deferal -> deferral
2013-03-27 23:07:43 +00:00
* This error only happens when NOWAIT is set, so deferral is disabled.
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
*/
if (error == ENOMEM)
return (error);
return (0);
}
int
bus_dmamap_load_mbuf(bus_dma_tag_t dmat, bus_dmamap_t map, struct mbuf *m0,
bus_dmamap_callback2_t *callback, void *callback_arg, int flags)
{
bus_dma_segment_t *segs;
int nsegs, error;
Move an assertion to the right spot; only bus_dmamap_load_mbuf(9) requires a pkthdr being present but that's not the case for either _bus_dmamap_load_mbuf_sg() or bus_dmamap_load_mbuf_sg(9). Reported by: sbruno MFC after: 1 week
2013-06-01 11:42:47 +00:00
M_ASSERTPKTHDR(m0);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
flags |= BUS_DMA_NOWAIT;
nsegs = -1;
error = _bus_dmamap_load_mbuf_sg(dmat, map, m0, NULL, &nsegs, flags);
++nsegs;
segs = _bus_dmamap_complete(dmat, map, NULL, nsegs, error);
if (error)
(*callback)(callback_arg, segs, 0, 0, error);
else
(*callback)(callback_arg, segs, nsegs, m0->m_pkthdr.len, error);
CTR5(KTR_BUSDMA, "%s: tag %p tag flags 0x%x error %d nsegs %d",
__func__, dmat, flags, error, nsegs);
return (error);
}
int
bus_dmamap_load_mbuf_sg(bus_dma_tag_t dmat, bus_dmamap_t map, struct mbuf *m0,
bus_dma_segment_t *segs, int *nsegs, int flags)
{
int error;
flags |= BUS_DMA_NOWAIT;
*nsegs = -1;
error = _bus_dmamap_load_mbuf_sg(dmat, map, m0, segs, nsegs, flags);
++*nsegs;
_bus_dmamap_complete(dmat, map, segs, *nsegs, error);
return (error);
}
int
bus_dmamap_load_uio(bus_dma_tag_t dmat, bus_dmamap_t map, struct uio *uio,
bus_dmamap_callback2_t *callback, void *callback_arg, int flags)
{
bus_dma_segment_t *segs;
int nsegs, error;
flags |= BUS_DMA_NOWAIT;
nsegs = -1;
error = _bus_dmamap_load_uio(dmat, map, uio, &nsegs, flags);
nsegs++;
segs = _bus_dmamap_complete(dmat, map, NULL, nsegs, error);
if (error)
(*callback)(callback_arg, segs, 0, 0, error);
else
(*callback)(callback_arg, segs, nsegs, uio->uio_resid, error);
CTR5(KTR_BUSDMA, "%s: tag %p tag flags 0x%x error %d nsegs %d",
Do not add 1 to nsegs before passing to CTR5(), since nsegs has already been incremented before these calls. Sponsored by: Intel
2013-03-29 15:54:12 +00:00
__func__, dmat, flags, error, nsegs);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
return (error);
}
int
bus_dmamap_load_ccb(bus_dma_tag_t dmat, bus_dmamap_t map, union ccb *ccb,
bus_dmamap_callback_t *callback, void *callback_arg,
int flags)
{
bus_dma_segment_t *segs;
struct ccb_hdr *ccb_h;
struct memdesc mem;
int error;
int nsegs;
ccb_h = &ccb->ccb_h;
if ((ccb_h->flags & CAM_DIR_MASK) == CAM_DIR_NONE) {
callback(callback_arg, NULL, 0, 0);
return (0);
}
if ((flags & BUS_DMA_NOWAIT) == 0) {
mem = memdesc_ccb(ccb);
_bus_dmamap_waitok(dmat, map, &mem, callback, callback_arg);
}
nsegs = -1;
error = _bus_dmamap_load_ccb(dmat, map, ccb, &nsegs, flags);
nsegs++;
Add CTR5() to bus_dmamap_load_ccb, similar to other bus_dmamap_load_* functions. Sponsored by: Intel
2013-03-29 16:00:16 +00:00
CTR5(KTR_BUSDMA, "%s: tag %p tag flags 0x%x error %d nsegs %d",
__func__, dmat, flags, error, nsegs);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
if (error == EINPROGRESS)
return (error);
segs = _bus_dmamap_complete(dmat, map, NULL, nsegs, error);
if (error)
(*callback)(callback_arg, segs, 0, error);
else
(*callback)(callback_arg, segs, nsegs, error);
/*
* Return ENOMEM to the caller so that it can pass it up the stack.
Add bus_dmamap_load_bio for non-CAM disk drivers that wish to enable unmapped I/O. Sponsored by: Intel Reviewed by: kib
2013-03-29 16:26:25 +00:00
* This error only happens when NOWAIT is set, so deferral is disabled.
*/
if (error == ENOMEM)
return (error);
return (0);
}
int
bus_dmamap_load_bio(bus_dma_tag_t dmat, bus_dmamap_t map, struct bio *bio,
bus_dmamap_callback_t *callback, void *callback_arg,
int flags)
{
bus_dma_segment_t *segs;
struct memdesc mem;
int error;
int nsegs;
if ((flags & BUS_DMA_NOWAIT) == 0) {
mem = memdesc_bio(bio);
_bus_dmamap_waitok(dmat, map, &mem, callback, callback_arg);
}
nsegs = -1;
error = _bus_dmamap_load_bio(dmat, map, bio, &nsegs, flags);
nsegs++;
CTR5(KTR_BUSDMA, "%s: tag %p tag flags 0x%x error %d nsegs %d",
__func__, dmat, flags, error, nsegs);
if (error == EINPROGRESS)
return (error);
segs = _bus_dmamap_complete(dmat, map, NULL, nsegs, error);
if (error)
(*callback)(callback_arg, segs, 0, error);
else
(*callback)(callback_arg, segs, nsegs, error);
/*
* Return ENOMEM to the caller so that it can pass it up the stack.
deferal -> deferral
2013-03-27 23:07:43 +00:00
* This error only happens when NOWAIT is set, so deferral is disabled.
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
*/
if (error == ENOMEM)
return (error);
return (0);
}
int
bus_dmamap_load_mem(bus_dma_tag_t dmat, bus_dmamap_t map,
struct memdesc *mem, bus_dmamap_callback_t *callback,
void *callback_arg, int flags)
{
bus_dma_segment_t *segs;
int error;
int nsegs;
if ((flags & BUS_DMA_NOWAIT) == 0)
_bus_dmamap_waitok(dmat, map, mem, callback, callback_arg);
nsegs = -1;
error = 0;
switch (mem->md_type) {
case MEMDESC_VADDR:
error = _bus_dmamap_load_buffer(dmat, map, mem->u.md_vaddr,
mem->md_opaque, kernel_pmap, flags, NULL, &nsegs);
break;
case MEMDESC_PADDR:
error = _bus_dmamap_load_phys(dmat, map, mem->u.md_paddr,
mem->md_opaque, flags, NULL, &nsegs);
break;
case MEMDESC_VLIST:
error = _bus_dmamap_load_vlist(dmat, map, mem->u.md_list,
Add asynchronous command support to the pass(4) driver, and the new camdd(8) utility. CCBs may be queued to the driver via the new CAMIOQUEUE ioctl, and completed CCBs may be retrieved via the CAMIOGET ioctl. User processes can use poll(2) or kevent(2) to get notification when I/O has completed. While the existing CAMIOCOMMAND blocking ioctl interface only supports user virtual data pointers in a CCB (generally only one per CCB), the new CAMIOQUEUE ioctl supports user virtual and physical address pointers, as well as user virtual and physical scatter/gather lists. This allows user applications to have more flexibility in their data handling operations. Kernel memory for data transferred via the queued interface is allocated from the zone allocator in MAXPHYS sized chunks, and user data is copied in and out. This is likely faster than the vmapbuf()/vunmapbuf() method used by the CAMIOCOMMAND ioctl in configurations with many processors (there are more TLB shootdowns caused by the mapping/unmapping operation) but may not be as fast as running with unmapped I/O. The new memory handling model for user requests also allows applications to send CCBs with request sizes that are larger than MAXPHYS. The pass(4) driver now limits queued requests to the I/O size listed by the SIM driver in the maxio field in the Path Inquiry (XPT_PATH_INQ) CCB. There are some things things would be good to add: 1. Come up with a way to do unmapped I/O on multiple buffers. Currently the unmapped I/O interface operates on a struct bio, which includes only one address and length. It would be nice to be able to send an unmapped scatter/gather list down to busdma. This would allow eliminating the copy we currently do for data. 2. Add an ioctl to list currently outstanding CCBs in the various queues. 3. Add an ioctl to cancel a request, or use the XPT_ABORT CCB to do that. 4. Test physical address support. Virtual pointers and scatter gather lists have been tested, but I have not yet tested physical addresses or scatter/gather lists. 5. Investigate multiple queue support. At the moment there is one queue of commands per pass(4) device. If multiple processes open the device, they will submit I/O into the same queue and get events for the same completions. This is probably the right model for most applications, but it is something that could be changed later on. Also, add a new utility, camdd(8) that uses the asynchronous pass(4) driver interface. This utility is intended to be a basic data transfer/copy utility, a simple benchmark utility, and an example of how to use the asynchronous pass(4) interface. It can copy data to and from pass(4) devices using any target queue depth, starting offset and blocksize for the input and ouptut devices. It currently only supports SCSI devices, but could be easily extended to support ATA devices. It can also copy data to and from regular files, block devices, tape devices, pipes, stdin, and stdout. It does not support queueing multiple commands to any of those targets, since it uses the standard read(2)/write(2)/writev(2)/readv(2) system calls. The I/O is done by two threads, one for the reader and one for the writer. The reader thread sends completed read requests to the writer thread in strictly sequential order, even if they complete out of order. That could be modified later on for random I/O patterns or slightly out of order I/O. camdd(8) uses kqueue(2)/kevent(2) to get I/O completion events from the pass(4) driver and also to send request notifications internally. For pass(4) devcies, camdd(8) uses a single buffer (CAM_DATA_VADDR) per CAM CCB on the reading side, and a scatter/gather list (CAM_DATA_SG) on the writing side. In addition to testing both interfaces, this makes any potential reblocking of I/O easier. No data is copied between the reader and the writer, but rather the reader's buffers are split into multiple I/O requests or combined into a single I/O request depending on the input and output blocksize. For the file I/O path, camdd(8) also uses a single buffer (read(2), write(2), pread(2) or pwrite(2)) on reads, and a scatter/gather list (readv(2), writev(2), preadv(2), pwritev(2)) on writes. Things that would be nice to do for camdd(8) eventually: 1. Add support for I/O pattern generation. Patterns like all zeros, all ones, LBA-based patterns, random patterns, etc. Right Now you can always use /dev/zero, /dev/random, etc. 2. Add support for a "sink" mode, so we do only reads with no writes. Right now, you can use /dev/null. 3. Add support for automatic queue depth probing, so that we can figure out the right queue depth on the input and output side for maximum throughput. At the moment it defaults to 6. 4. Add support for SATA device passthrough I/O. 5. Add support for random LBAs and/or lengths on the input and output sides. 6. Track average per-I/O latency and busy time. The busy time and latency could also feed in to the automatic queue depth determination. sys/cam/scsi/scsi_pass.h: Define two new ioctls, CAMIOQUEUE and CAMIOGET, that queue and fetch asynchronous CAM CCBs respectively. Although these ioctls do not have a declared argument, they both take a union ccb pointer. If we declare a size here, the ioctl code in sys/kern/sys_generic.c will malloc and free a buffer for either the CCB or the CCB pointer (depending on how it is declared). Since we have to keep a copy of the CCB (which is fairly large) anyway, having the ioctl malloc and free a CCB for each call is wasteful. sys/cam/scsi/scsi_pass.c: Add asynchronous CCB support. Add two new ioctls, CAMIOQUEUE and CAMIOGET. CAMIOQUEUE adds a CCB to the incoming queue. The CCB is executed immediately (and moved to the active queue) if it is an immediate CCB, but otherwise it will be executed in passstart() when a CCB is available from the transport layer. When CCBs are completed (because they are immediate or passdone() if they are queued), they are put on the done queue. If we get the final close on the device before all pending I/O is complete, all active I/O is moved to the abandoned queue and we increment the peripheral reference count so that the peripheral driver instance doesn't go away before all pending I/O is done. The new passcreatezone() function is called on the first call to the CAMIOQUEUE ioctl on a given device to allocate the UMA zones for I/O requests and S/G list buffers. This may be good to move off to a taskqueue at some point. The new passmemsetup() function allocates memory and scatter/gather lists to hold the user's data, and copies in any data that needs to be written. For virtual pointers (CAM_DATA_VADDR), the kernel buffer is malloced from the new pass(4) driver malloc bucket. For virtual scatter/gather lists (CAM_DATA_SG), buffers are allocated from a new per-pass(9) UMA zone in MAXPHYS-sized chunks. Physical pointers are passed in unchanged. We have support for up to 16 scatter/gather segments (for the user and kernel S/G lists) in the default struct pass_io_req, so requests with longer S/G lists require an extra kernel malloc. The new passcopysglist() function copies a user scatter/gather list to a kernel scatter/gather list. The number of elements in each list may be different, but (obviously) the amount of data stored has to be identical. The new passmemdone() function copies data out for the CAM_DATA_VADDR and CAM_DATA_SG cases. The new passiocleanup() function restores data pointers in user CCBs and frees memory. Add new functions to support kqueue(2)/kevent(2): passreadfilt() tells kevent whether or not the done queue is empty. passkqfilter() adds a knote to our list. passreadfiltdetach() removes a knote from our list. Add a new function, passpoll(), for poll(2)/select(2) to use. Add devstat(9) support for the queued CCB path. sys/cam/ata/ata_da.c: Add support for the BIO_VLIST bio type. sys/cam/cam_ccb.h: Add a new enumeration for the xflags field in the CCB header. (This doesn't change the CCB header, just adds an enumeration to use.) sys/cam/cam_xpt.c: Add a new function, xpt_setup_ccb_flags(), that allows specifying CCB flags. sys/cam/cam_xpt.h: Add a prototype for xpt_setup_ccb_flags(). sys/cam/scsi/scsi_da.c: Add support for BIO_VLIST. sys/dev/md/md.c: Add BIO_VLIST support to md(4). sys/geom/geom_disk.c: Add BIO_VLIST support to the GEOM disk class. Re-factor the I/O size limiting code in g_disk_start() a bit. sys/kern/subr_bus_dma.c: Change _bus_dmamap_load_vlist() to take a starting offset and length. Add a new function, _bus_dmamap_load_pages(), that will load a list of physical pages starting at an offset. Update _bus_dmamap_load_bio() to allow loading BIO_VLIST bios. Allow unmapped I/O to start at an offset. sys/kern/subr_uio.c: Add two new functions, physcopyin_vlist() and physcopyout_vlist(). sys/pc98/include/bus.h: Guard kernel-only parts of the pc98 machine/bus.h header with #ifdef _KERNEL. This allows userland programs to include <machine/bus.h> to get the definition of bus_addr_t and bus_size_t. sys/sys/bio.h: Add a new bio flag, BIO_VLIST. sys/sys/uio.h: Add prototypes for physcopyin_vlist() and physcopyout_vlist(). share/man/man4/pass.4: Document the CAMIOQUEUE and CAMIOGET ioctls. usr.sbin/Makefile: Add camdd. usr.sbin/camdd/Makefile: Add a makefile for camdd(8). usr.sbin/camdd/camdd.8: Man page for camdd(8). usr.sbin/camdd/camdd.c: The new camdd(8) utility. Sponsored by: Spectra Logic MFC after: 1 week
2015-12-03 20:54:55 +00:00
mem->md_opaque, kernel_pmap, &nsegs, flags, 0, SIZE_T_MAX);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
break;
case MEMDESC_PLIST:
error = _bus_dmamap_load_plist(dmat, map, mem->u.md_list,
mem->md_opaque, &nsegs, flags);
break;
case MEMDESC_BIO:
error = _bus_dmamap_load_bio(dmat, map, mem->u.md_bio,
&nsegs, flags);
break;
case MEMDESC_UIO:
error = _bus_dmamap_load_uio(dmat, map, mem->u.md_uio,
&nsegs, flags);
break;
case MEMDESC_MBUF:
error = _bus_dmamap_load_mbuf_sg(dmat, map, mem->u.md_mbuf,
NULL, &nsegs, flags);
break;
case MEMDESC_CCB:
error = _bus_dmamap_load_ccb(dmat, map, mem->u.md_ccb, &nsegs,
flags);
break;
}
nsegs++;
CTR5(KTR_BUSDMA, "%s: tag %p tag flags 0x%x error %d nsegs %d",
Do not add 1 to nsegs before passing to CTR5(), since nsegs has already been incremented before these calls. Sponsored by: Intel
2013-03-29 15:54:12 +00:00
__func__, dmat, flags, error, nsegs);
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
if (error == EINPROGRESS)
return (error);
segs = _bus_dmamap_complete(dmat, map, NULL, nsegs, error);
if (error)
(*callback)(callback_arg, segs, 0, error);
else
(*callback)(callback_arg, segs, nsegs, 0);
/*
* Return ENOMEM to the caller so that it can pass it up the stack.
deferal -> deferral
2013-03-27 23:07:43 +00:00
* This error only happens when NOWAIT is set, so deferral is disabled.
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
*/
if (error == ENOMEM)
return (error);
return (0);
}
Reference in New Issue Copy Permalink