- 09 May, 2007 2 commits
-
-
Roland Dreier authored
When memory pinned with ib_umem_get() is released, ib_umem_release() needs to subtract the amount of memory being unpinned from mm->locked_vm. However, ib_umem_release() may be called with mm->mmap_sem already held for writing if the memory is being released as part of an munmap() call, so it is sometimes necessary to defer this accounting into a workqueue. However, the work struct used to defer this accounting is dynamically allocated before it is queued, so there is the possibility of failing that allocation. If the allocation fails, then ib_umem_release has no choice except to bail out and leave the process with a permanently elevated locked_vm. Fix this by allocating the structure to defer accounting as part of the original struct ib_umem, so there's no possibility of failing a later allocation if creating the struct ib_umem and pinning memory succeeds. Signed-off-by:
Roland Dreier <rolandd@cisco.com>
-
Roland Dreier authored
Export ib_umem_get()/ib_umem_release() and put low-level drivers in control of when to call ib_umem_get() to pin and DMA map userspace, rather than always calling it in ib_uverbs_reg_mr() before calling the low-level driver's reg_user_mr method. Also move these functions to be in the ib_core module instead of ib_uverbs, so that driver modules using them do not depend on ib_uverbs. This has a number of advantages: - It is better design from the standpoint of making generic code a library that can be used or overridden by device-specific code as the details of specific devices dictate. - Drivers that do not need to pin userspace memory regions do not need to take the performance hit of calling ib_mem_get(). For example, although I have not tried to implement it in this patch, the ipath driver should be able to avoid pinning memory and just use copy_{to,from}_user() to access userspace memory regions. - Buffers that need special mapping treatment can be identified by the low-level driver. For example, it may be possible to solve some Altix-specific memory ordering issues with mthca CQs in userspace by mapping CQ buffers with extra flags. - Drivers that need to pin and DMA map userspace memory for things other than memory regions can use ib_umem_get() directly, instead of hacks using extra parameters to their reg_phys_mr method. For example, the mlx4 driver that is pending being merged needs to pin and DMA map QP and CQ buffers, but it does not need to create a memory key for these buffers. So the cleanest solution is for mlx4 to call ib_umem_get() in the create_qp and create_cq methods. Signed-off-by:
Roland Dreier <rolandd@cisco.com>
-
- 08 May, 2007 38 commits
-
-
git://jdelvare.pck.nerim.net/jdelvare-2.6Linus Torvalds authored
* 'hwmon-for-linus' of git://jdelvare.pck.nerim.net/jdelvare-2.6: (32 commits) Use menuconfig objects - hwmon hwmon/smsc47b397: Use dynamic sysfs callbacks hwmon/smsc47b397: Convert to a platform driver hwmon/w83781d: Deprecate W83627HF support hwmon/w83781d: Use dynamic sysfs callbacks hwmon/w83781d: Be less i2c_client-centric hwmon/w83781d: Clean up conversion macros hwmon/w83781d: No longer use i2c-isa hwmon/ams: Do not print error on systems without apple motion sensor hwmon/ams: Fix I2C read retry logic hwmon: New AD7416, AD7417 and AD7418 driver hwmon/coretemp: Add documentation hwmon: New coretemp driver i386: Use functions from library in msr driver i386: Add safe variants of rdmsr_on_cpu and wrmsr_on_cpu hwmon/lm75: Use dynamic sysfs callbacks hwmon/lm78: Use dynamic sysfs callbacks hwmon/lm78: Be less i2c_client-centric hwmon/lm78: No longer use i2c-isa hwmon: New max6650 driver ...
-
Geert Uytterhoeven authored
Commit c5a4dd8b introduced the following compiler warnings: net/sunrpc/sched.c:766: warning: format '%u' expects type 'unsigned int', but argument 3 has type 'size_t' net/sunrpc/sched.c:785: warning: format '%u' expects type 'unsigned int', but argument 2 has type 'size_t' - Use %zu to format size_t - Kill 2 useless casts Signed-off-by:
Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
git://oss.sgi.com:8090/xfs/xfs-2.6Linus Torvalds authored
* git://oss.sgi.com:8090/xfs/xfs-2.6: [XFS] Add lockdep support for XFS [XFS] Fix race in xfs_write() b/w dmapi callout and direct I/O checks. [XFS] Get rid of redundant "required" in msg. [XFS] Export via a function xfs_buftarg_list for use by kdb/xfsidbg. [XFS] Remove unused ilen variable and references. [XFS] Fix to prevent the notorious 'NULL files' problem after a crash. [XFS] Fix race condition in xfs_write(). [XFS] Fix uquota and oquota enforcement problems. [XFS] propogate return codes from flush routines [XFS] Fix quotaon syscall failures for group enforcement requests. [XFS] Invalidate quotacheck when mounting without a quota type. [XFS] reducing the number of random number functions. [XFS] remove more misc. unused args [XFS] the "aendp" arg to xfs_dir2_data_freescan is always NULL, remove it. [XFS] The last argument "lsn" of xfs_trans_commit() is always called with
-
Linus Torvalds authored
* 'reset-seq' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev: [libata reset-seq] build and merge fixes libata: reimplement reset sequencing libata: improve ata_std_prereset() libata: improve 0xff status handling libata: add deadline support to prereset and reset methods
-
Linus Torvalds authored
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6: (40 commits) [netdrvr] atl1: fix build pasemi_mac: Use local-mac-address instead of mac-address if available pasemi_mac: PHY support pasemi_mac: Add msglevel support and "debug" module param pasemi_mac: Logic cleanup / rx performance improvements pasemi_mac: Minor cleanup / define fixes pasemi_mac: Add SKB reuse / copy-break pasemi_mac: Timer and interrupt fixes pasemi_mac: Abstract and fix up interrupt restart routines pasemi_mac: Move the IRQ mapping from the PCI layer to the driver tc35815: Remove unnecessary skb->dev assignment drivers/net/dm9000: Convert to generic boolean AT91RM9200 Ethernet: Fix multicast addressing AT91RM9200 Ethernet: Support additional PHYs PCMCIA-NETDEV : xirc2ps_cs: bugfix of multicast code sky2: re-enable 88E8056 for most motherboards MIPS: Drop unnecessary CONFIG_ISA from RBTX49XX ne: MIPS: Use platform_driver for ne on RBTX49XX ne: Add NEEDS_PORTLIST to control ISA auto-probe ne: Misc fixes for platform driver. ... Fix conflict in drivers/net/pasemi_mac.c (get_property() got renamed to of_get_property()) manually. Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Linus Torvalds authored
* 'drm-patches' of master.kernel.org:/pub/scm/linux/kernel/git/airlied/drm-2.6: via: Make sure we flush write-combining using a follow-up read. via: Try to improve command-buffer chaining. drm: remove old taskqueue remnant drm: rename badly named define and cleanup ioctl code spacing radeon: Don't mess up page flipping when a file descriptor is closed. drm/radeon: upgrade to 1.27 - make PCI GART more flexible
-
Linus Torvalds authored
* master.kernel.org:/pub/scm/linux/kernel/git/dtor/input: Input: move USB miscellaneous devices under drivers/input/misc Input: move USB mice under drivers/input/mouse Input: move USB gamepads under drivers/input/joystick Input: move USB touchscreens under drivers/input/touchscreen Input: move USB tablets under drivers/input/tablet Input: i8042 - fix AUX port detection with some chips Input: aaed2000_kbd - convert to use polldev library Input: drivers/usb/input - usb_buffer_free() cleanup Input: synaptics - don't complain about failed resets Input: pull input.h into uinpit.h Input: drivers/usb/input - fix sparse warnings (signedness) Input: evdev - fix some sparse warnings (signedness, shadowing) Input: drivers/joystick - fix various sparse warnings Input: force feedback - make sure effect is present before playing
-
git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpcLinus Torvalds authored
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc: (77 commits) [POWERPC] Abolish powerpc_flash_init() [POWERPC] Early serial debug support for PPC44x [POWERPC] Support for the Ebony 440GP reference board in arch/powerpc [POWERPC] Add device tree for Ebony [POWERPC] Add powerpc/platforms/44x, disable platforms/4xx for now [POWERPC] MPIC U3/U4 MSI backend [POWERPC] MPIC MSI allocator [POWERPC] Enable MSI mappings for MPIC [POWERPC] Tell Phyp we support MSI [POWERPC] RTAS MSI implementation [POWERPC] PowerPC MSI infrastructure [POWERPC] Rip out the existing powerpc msi stubs [POWERPC] Remove use of 4level-fixup.h for ppc32 [POWERPC] Add powerpc PCI-E reset API implementation [POWERPC] Holly bootwrapper [POWERPC] Holly DTS [POWERPC] Holly defconfig [POWERPC] Add support for 750CL Holly board [POWERPC] Generalize tsi108 PCI setup [POWERPC] Generalize tsi108 PHY types ... Fixed conflict in include/asm-powerpc/kdebug.h manually Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Roland McGrath authored
Signed-off-by:
Roland McGrath <roland@redhat.com> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
git://git.kernel.dk/data/git/linux-2.6-blockLinus Torvalds authored
* 'for-2.6.22' of git://git.kernel.dk/data/git/linux-2.6-block: [PATCH] ll_rw_blk: fix missing bounce in blk_rq_map_kern() [PATCH] splice: always call into page_cache_readahead() [PATCH] splice(): fix interaction with readahead
-
git://git.kernel.org/pub/scm/linux/kernel/git/shaggy/jfs-2.6Linus Torvalds authored
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/shaggy/jfs-2.6: JFS: Fix race waking up jfsIO kernel thread JFS: use __set_current_state() Copy i_flags to jfs inode flags on write JFS: document uid, gid, and umask mount options in jfs.txt
-
Andrew Morton authored
drivers/video/sm501fb.c: In function 'sm501fb_cursor': drivers/video/sm501fb.c:992: warning: format '%08x' expects type 'unsigned int', but argument 3 has type 'long unsigned int' drivers/video/sm501fb.c:992: warning: format '%08x' expects type 'unsigned int', but argument 4 has type 'long unsigned int' Cc: "Antonino A. Daplas" <adaplas@pol.net> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Andrei Konovalov authored
Add support for the video controller IP block included into Xilinx ML300 and ML403 reference designs. Signed-off-by:
Andrei Konovalov <akonovalov@ru.mvista.com> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
This patch fixes incorrect targets of jumps when an error occurs in the pm2fb_probe. Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
This patch adds a macro to register PCI ids table and corrects type of xxxfb_fix variable to avoid modpost warnings. Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
This patch removes two redundant fields in the pm2fb_par structure. Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
More corrections to skeletonfb.c file: - fixed types of more arguments - removed returned values in void functions - removed not existing fb_poll function - fixed closing comment typo - corrected misleading constant name ADDR Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Acked-By:
James Simmons <jsimmons@infradead.org> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Antonino A. Daplas authored
vga16fb does not just support 8-pixel wide rectangles, but rectangles with widths in multiples of 8. Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Antonino A. Daplas authored
Reported by James Pearson as: boot to run level 3 if not root, then make sure /dev/console is writeable login and type: setterm -blank 0 start X type into an xterm: while true; do echo "" > /dev/console; usleep 100000; done while the above loop is running switch to the text console and back again (Ctrl-Alt-F1 then Ctrl-Alt-F7) ... and the screen will be shifting (and wrapping) to the left. This problem stems from continuously writing text to the system console (which is in KD_TEXT mode) while the foreground console is in KD_GRAPHICS mode. Somewhere along the way, console printing got confused and omitted the KD_GRAPHICS/KD_TEXT test. Thus, vgacon attempted to scroll the screen of X, which causes X to shift. Fix by disallowing vgacon to touch the hardware when the vc is in KD_GRAPHICS mode. A definitive fix entails a full audit of the console code. Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Jan Engelhardt authored
Change Kconfig objects from "menu, config" into "menuconfig" so that the user can disable the whole feature without having to enter the menu first. Signed-off-by:
Jan Engelhardt <jengelh@gmx.de> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
This is a port of accelerated functions from 2.4 kernel. Only fillrect and copyarea are accelerated. Fillrect is not accelerated in 24-bit mode. [adaplas] Add appropriate flags Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Ondrej Zajicek authored
Move s3fb_get_tilemax to svgalib.c as svga_get_tilemax, because it reports limitation of other code from svgalib (svga_settile, svga_tilecopy, ...) Limit font width to 8 pixels in 4 bpp mode. Signed-off-by:
Ondrej Zajicek <santiago@crfreenet.org> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
This patch disables a memory clock setting if a board has been initialized by BIOS. This allows using the memory clock set by manufacturer of the board. This patch also sets default clock for 3dlabs Permedia 2V reference board's clock to 75MHz (BIOS setting for EONtronic Permedia 2Vboard), because the default 83MHz can be too high. Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
krzysztof.h1@wp.pl authored
This patch resets transparency settings when depth changes. Otherwise the 16 and 24-bit modes work incorrectly after switching from 32-bit mode. Signed-off-by: Krzysztof Helt < krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Antonino A. Daplas authored
- Check the console-to-fb mapping in fbcon_get_requirement(), otherwise the value returned may not be valid for the driver. - Minor cleanup Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Antonino A. Daplas authored
Clean up exit patch of fb_set_var(): - consolidate all return values into a single local variable - ensure that return values are valid error codes - fix fb_set_var() returning success when fb_check_caps() failed Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Adrian Bunk authored
Signed-off-by:
Adrian Bunk <bunk@stusta.de> Cc: Thomas Winischhofer <thomas@winischhofer.net> Cc: "Antonino A. Daplas" <adaplas@pol.net> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Matthias Kaehlcke authored
The virtual console driver uses a semaphore as mutex. Use the mutex API instead of the (binary) semaphore. Signed-off-by:
Matthias Kaehlcke <matthias.kaehlcke@gmail.com> Cc: "Antonino A. Daplas" <adaplas@pol.net> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Ville Syrjala authored
Laptops with Rage Mobility and 32bit memory interface seem to require halved XCLK to operate correctly. Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Ville Syrjala authored
Reorganize atyfb clock init code so command line clock overrides are effective for all chips. The old code would silently ignore some of the command line clock overrides with some chips. Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Ville Syrjala authored
Wait 5 ms instead of 500 us for the SPLL to lock. This matches the recommendation in mach64 programmer's guide. Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Antonino A. Daplas authored
Fix return value of nvidiafb_open(). Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Ondrej Zajicek authored
1) sets 'palette access disabled' during read from AR10. This is usually documented as needed for access AR01-AR0F, but on ARK Logic card it is needed for AR10 (otherwise read returns some nonsence and save_vga_text() returns before do anything). 2) do not restore AR10, because it was not changed during 3) remove modification of misc reg: /* force graphics mode */ vga_w(state->vgabase, VGA_MIS_W, misc | 1); as comment is misleading - LSB of misc reg does not set/reset graphics mode, but set color/mono adresses of CRT and some other regs. but these regs are not used during save/restore fonts. (it worked even when (misc | 1) was replaced by (misc & ~1) ). Signed-off-by:
Ondrej Zajicek <santiago@crfreenet.org> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Antonino A. Daplas authored
Implement fb_get_caps(). This will allow the driver to tell upper layer that it can only support 256 8x16 bitmaps when in tileblitting mode (var->bits_per_pixel == 0); Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Antonino A. Daplas authored
Check if the mode can properly display the screen. This will be needed by drivers where the capability is not constant with each mode. The function fb_set_var() will query fbcon the requirement, then it will query the driver (via a new hook fb_get_caps()) its capability. If the driver's capability cannot handle fbcon's requirement, then fb_set_var() will fail. For example, if a particular driver supports 2 modes where: mode1 = can only display 8x16 bitmaps mode2 = can display any bitmap then if current mode = mode2 and current font = 12x22 fbset <mode1> /* mode1 cannot handle 12x22 */ fbset will fail Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Dave Jones authored
If the chip locks up, we get into a long polling loop, where the softlockup detector kicks in. See https://bugzilla.redhat.com/bugzilla/attachment.cgi?id=151878 for an example. [adaplas] Chip lockup can occur at 3 points (flush, sync, and wait). Consolidate and allow the driver to go to safe mode cleanly. Signed-off-by:
Dave Jones <davej@redhat.com> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
This patch adds restrictions to calculations of m, n and p factors which sets the pixclock for the Permedia 2V. I found during tests that synchronization is unstable if m (divisor) is bigger than half of the n (numerator). The patch disallows such settings combination. Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Krzysztof Helt authored
Permedia 2V uses its own registers to set a memory clock. The patch adds these registers and uses them in the set_memclock() function. Signed-off-by:
Krzysztof Helt <krzysztof.h1@wp.pl> Signed-off-by:
Antonino Daplas <adaplas@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-