summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* While certain supported Symbios/LSI SCSI chips (532c896, 53c1000, 53c1010)se2009-10-311-1/+1
| | | | | | | | | | | | | | | | | | do support 64bit addresses, the current SCRIPTS code supports only 32bit addresses causing data corruption for buffer addresses >4GB. This problem affects 64bit machines with more than 4GB RAM or amd64 with 4GB and memory hole remapping. Work-around this problem with a bus_dma tag that requests bounce-buffers for addresses >4GB. This causes some overhead, but given the maximum SCSI bus speed of 160MB/s compared, the effect should hardly be noticeable. The problem was reported by Mike Watters (mike at mwatters net) who also verified that this fix cures the problem. Since this change is a NOOP on systems with less than 4GB RAM and fixes data corruption (in RAM and on disk) on systems with more than 4GB, I hope that this change is accepted for 8.0. MFC after: 3 days (pending approval)
* Allow newly added controllers to use full I/O sizes.mav2009-10-311-1/+3
|
* MFp4:mav2009-10-3114-155/+138
| | | | | | | | - Remove most of direct relations between ATA(4) peripherial and controller levels. It makes logic more transparent and is a mandatory step to wrap ATA(4) controller level into ATA-native CAM SIM. - Tune AHCI and SATA2 SiI drivers memory allocation a bit to allow bigger I/O transaction sizes without additional cost.
* The majors file was removed long ago, 0 should be used instead.brueffer2009-10-311-4/+2
| | | | | | PR: 139230 Submitted by: pluknet <pluknet@gmail.com> MFC after: 1 week
* MFp4:mav2009-10-311-40/+57
| | | | | Sync connection speed reporting with kernel. Report speed in identify command, same as done by inquiry.
* MFp4:mav2009-10-3110-517/+452
| | | | | | | | | | | | - Reduce code duplication in ATA XPT and PMP driver. - Move PIO size setting from ada driver to ATA XPT. It is XPT business to negotiate transfer details. ada driver is now stateless. - Report PIO size to SIM. It is required for correct PATA SIM operation. - Tune PMP scan timings. It workarounds some problems with SiI. - If reset hapens during PMP initialization - restart it. - Introduce early-initialized periph drivers, which are used during initial scan process. Use it for xpt, probe, aprobe and pmp. It gives pmp chance to finish scan before mountroot and numerate devices in right order.
* Unobfuscate unit number handling in apm(4).ed2009-10-311-13/+9
| | | | | | | | There is no need to use the lower 4 bits of the unit number to store the device type number. Just use 0 and 1 to distinguish them. devfs also guarantees that there can never be an open call on a device that has a unit number different to 0 and 1, so there is no need to check for this in open().
* Turn unused structure fields of cdevsw into spares.ed2009-10-311-6/+4
| | | | | | | | | | | | d_uid, d_gid and d_mode are unused, because permissions are stored in cdevpriv nowadays. d_kind doesn't seem to be used at all. We no longer keep a list of cdevsw's, so d_list is also unused. uid_t and gid_t are 32 bits, but mode_t is 16 bits, Because of alignment constraints of d_kind, we can safely turn it into three 32-bit integers. d_kind and d_list is equal in size to three pointers. Discussed with: kib
* MFp4:mav2009-10-311-1/+28
| | | | | | Ensure target/lun passed from user-level supported on this bus. Scanning unsupported IDs causes different issues from duplicate devices to system crash.
* - zfs_zaccess() can handle VAPPEND too, so map V_APPEND to VAPPEND and callpjd2009-10-303-10/+23
| | | | | | | | | | | | | zfs_access() instead of vaccess() in this case as well. - If VADMIN is specified with another V* flag (unlikely) call both zfs_access() and vaccess() after spliting V* flags. This fixes "dirtying snapshot!" panic. PR: kern/139806 Reported by: Carl Chave <carl@chave.us> In co-operation with: jh MFC after: 3 days
* When reading input from a file or device (via -f option) set themckusick2009-10-301-0/+1
| | | | | | input to be in non-buffering mode so that input lines are logged as they occur rather than being saved up until a buffer's worth of input has been logged.
* Add support for different request block format used by Gen-IIe Marvell SATA.mav2009-10-303-30/+70
| | | | This adds support for Marvell 6042/7042 chips and Adaptec 1430SA controller.
* Fix botched git -> svn merge.rnoland2009-10-301-1/+1
| | | | MFC after: 2 weeks
* Cleanup in r600_blitrnoland2009-10-301-29/+39
| | | | | | | | | - Don't bother to assign vb until we know we have enough space - Add variables for sx2, sy2, dx2, dy2 so that these aren't calculated over and over, also reduce chance of errors. - Use switch to assign color/format MFC after: 3 days
* A bit of cleanup work on radeon_freelist_get()rnoland2009-10-301-41/+4
| | | | | | | * Fix the main loop to search all buffers before sleeping. * Remove dead code MFC after: 3 days
* Some general cleanup of scatter/gather memory allocationrnoland2009-10-301-44/+25
| | | | | | | | | - We don't need to check malloc return values with M_WAITOK - remove variables that we don't really need - cleanup the error paths by just calling drm_sg_cleanup() - fix drm_sg_cleanup() to be safe to call at any time MFC after: 2 weeks
* Use system specified memory barriers rather than rolling our own.rnoland2009-10-301-17/+3
|
* Fix blitter support for RS880 chipsrnoland2009-10-301-3/+3
| | | | MFC after: 3 days
* Add support for Adaptec 39320LPE adapters.brueffer2009-10-301-0/+7
| | | | | | | PR: 124202 Submitted by: Andre Albsmeier <Andre.Albsmeier@siemens.com> Reviewed by: gibbs MFC after: 1 week
* Make procstat -k work on PowerPC by avoiding mistakenly using signednwhitehorn2009-10-301-2/+2
| | | | | compares with a low address (0x1000) and a high address (the KVA kernel stack).
* Add notes pointing out that bsdiff does not store file hashes andcperciva2009-10-302-0/+40
| | | | | | | | | bspatch thus does not verify file hashes, and that consequently it is recommended that users store hashes separately and verify files before and after running bspatch. Requested by: BugMagnet MFC after: 1 week
* Trapsignal() and postsig() call kern_sigprocmask() with both processkib2009-10-302-20/+17
| | | | | | | | | | | | lock and curproc->p_sigacts->ps_mtx. Reschedule_signals may need to have ps_mtx locked to decide and wakeup a thread, causing recursion on the mutex. Inform kern_sigprocmask() and reschedule_signals() about lock state of the ps_mtx by new flag SIGPROCMASK_PS_LOCKED to avoid recursion. Reported and tested by: keramida MFC after: 1 month
* Add missing ATA kernel options dependencies.mav2009-10-291-2/+2
| | | | MFC after: 3 days
* - Numerous whitespace and style fixes.jhb2009-10-291-127/+138
| | | | | - More descriptive error messages when failing to parse components during attach.
* Improve round robin stream scheduler and cleanup some code.tuexen2009-10-291-21/+24
| | | | | Approved by: rrs (mentor) MFC after: 3 days
* When fetching sum stats (vmstat -s) from a crash dump, fetch per-CPU countsjhb2009-10-291-0/+81
| | | | | | | | and sum them to form the total counts. PR: bin/135893 Submitted by: Mikolaj Golub to my trociny of gmail MFC after: 1 week
* Add additional featuresState.fBits entries to simplify compiling andemaste2009-10-291-1/+4
| | | | | | testing Adaptec's vendor driver. Submitted by: Adaptec, driver 17517
* Rename aac_srb32 to aac_srb, to match Adaptec's vendor driver.emaste2009-10-292-4/+4
|
* Revert part of r198363, there is no "device cam", it isbrueffer2009-10-292-2/+0
| | | | | | included in "device scbus". MFC after: 3 days
* Rename aac_fast_intr to aac_filter to reflect its current use. Eliminateemaste2009-10-292-12/+8
| | | | | | | the fallback of using the filter as an interrupt handler, as it is no longer needed. Discussed with: scottl, jhb
* Trapsignal() calls kern_sigprocmask() when delivering catched signalkib2009-10-291-1/+2
| | | | | | | with proc lock held. Reported and tested by: Mykola Dzham freebsd at levsha org ua MFC after: 1 month
* Turn off Altivec data-stream prefetching before going into power-savenwhitehorn2009-10-291-3/+21
| | | | mode on those CPUs that need it.
* Define identify fields described in CF specification.mav2009-10-291-1/+5
|
* Include the output of the ddb(4) capture buffer.jhb2009-10-291-0/+7
| | | | | Submitted by: Mikolaj Golub to my trociny of gmail MFC after: 3 days
* When extracting the capture buffer from a crashdump, only read the validjhb2009-10-291-10/+10
| | | | | | | | | | portion of the capture buffer (db_capture_bufoff vs db_capture_bufsize). This could result in outputting garbage (e.g. lots of 'p' characters if DIAGNOSTIC is enabled) after the end of the capture buffer. While here, fix a spelling nit. Reported by: Mikolaj Golub to my trociny of gmail MFC after: 3 days
* Add some magic taken from OS X and Linux to support early revision K2nwhitehorn2009-10-291-0/+31
| | | | | | SATA controllers, like those found on the G5 Xserve. Reviewed by: mav
* Turn off use of ATA_A_4BIT on modern hardware. This flag was alreadynwhitehorn2009-10-291-1/+10
| | | | | | | | | | obsoleted in 1996 by ATA-2, and crashes some modern hardware like some revisions of the Serverworks K2 SATA controller. Even very ancient hardware seems not to require it. In the unlikely event this causes problems, the previous behavior can be re-enabled by defining ATA_LEGACY_SUPPORT at the top of this file. Reviewed by: Alexander Motin <mav@freebsd.org>
* Update the route's sequence number upon receiving a RANN.rpaulo2009-10-291-9/+12
| | | | MFC after: 3 days
* Fix style issue.kib2009-10-291-1/+1
|
* Update some comments regarding ktr(4).ru2009-10-291-6/+5
|
* Back in 2003, get_cyclecount() was changed to use binuptime() insteadru2009-10-291-10/+4
| | | | | | of nanotime(). Reflect this change in a manpage. Reviewed by: phk, markm
* HZ is now 1000 on most platforms, update a comment.ru2009-10-291-1/+1
| | | | Reviewed by: phk, markm
* Fix some problems with effective mmap() offsets > 32 bits. This wasjhb2009-10-282-70/+61
| | | | | | | | | | | | partially fixed on amd64 earlier. Rather than forcing linux_mmap_common() to use a 32-bit offset, have it accept a 64-bit file offset. This offset is then passed to the real mmap() call. Rather than inventing a structure to hold the normal linux_mmap args that has a 64-bit offset, just pass each of the arguments individually to linux_mmap_common() since that more closes matches the existing style of various kern_foo() functions. Submitted by: Christian Zander @ Nvidia MFC after: 1 week
* Remove spurious `)`brueffer2009-10-281-1/+1
| | | | | | PR: 137758 Submitted by: Henning Petersen <henning.petersen@t-online.de> MFC after: 1 week
* Fix date (1) and SEE ALSO section.brueffer2009-10-281-3/+3
| | | | | Submitted by: Ulrich Spoerlein (1) MFC after: 1 week
* Initialize f_rabuf in the raw device case. A subsequent close()brueffer2009-10-281-0/+1
| | | | | | | | later on would try to free it, leading to a crash. PR: 93998 Submitted by: neel MFC after: 1 week
* Do first controller time sync after 1 minute, as in Adaptec's vendoremaste2009-10-281-1/+1
| | | | driver.
* Close a stream file descriptor leak.brueffer2009-10-281-0/+1
| | | | | | PR: 138130 Submitted by: Patroklos Argyroudis <argp@census-labs.com> MFC after: 1 week
* Move pselect(3) man page to section 2.kib2009-10-284-5/+8
| | | | | Noted by: jhb MFC after: 1 month
* Close a file descriptor leak in an error case.brueffer2009-10-281-0/+1
| | | | | | PR: 138374 Submitted by: Patroklos Argyroudis <argp@census-labs.com> MFC after: 1 week
OpenPOWER on IntegriCloud