- 06 Sep, 2011 7 commits
-
-
Thomas Hellstrom authored
Will be needed for queries and drm event-driven throttling. As a benefit, they help avoid stale user-space fence handles. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
Break out on-demand enabling and disabling of fence irqs to make the function more readable. Also make dev_priv->fence_queue_waiters an int instead of an atomic_t since we only manipulate it with dev_priv->hw_mutex held. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
This is needed before we introduce the fence objects. Otherwise this will be even more confusing. The plan is to use the following: seqno: A 32-bit sequence number that may be passed in the fifo. marker: Objects, carrying a seqno, that track fifo submission time. They are used for fifo lag based throttling. fence objects: Kernel space objects, possibly accessible from user-space and carrying a 32-bit seqno together with signaled status. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
Since we don't allow user-space to map the fifo anymore, add a parameter to get fifo hw version and an ioctl to copy the 3D capabilities. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecranz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
This was previously used by user-space to check whether a fence sequence had passed or not. With fence objects that's not needed anymore. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
It doesn't seem like its needed. If this turns out to be an incorrect assumption, we can reinstate it. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
It was only used for bringup debugging, and probably doesn't work anymore. Remove it. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
- 01 Sep, 2011 10 commits
-
-
Thomas Hellstrom authored
Guest Memory Regions 2 is a way to bind pages to the GPU, but using the FIFO instead of an io-submitted descriptor chain. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: José Fonseca <jfonseca@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
When GMR2 is available, make sure we restrict the number of used GMR pages to the limit indicated by the device. This is done by failing a GMRID allocation if the total number of GMR pages exceeds the limit. As a result TTM will then start evicting buffers in GMR memory on a LRU basis until the allocation succeeds. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
Previously this was not done when any 3D resource was active, since that meant disabling the fifo with all 3D state lost. Now, if there are still 3D resources active, we use the svga hide feature. This fixes X server VT switching with 3D enabled. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
Perform all command stream validation in a bounce buffer separate from the fifo. This makes the fifo available to all validation-generated commands, which would otherwise attempt to grab the fifo recursively, causing a deadlock. This is in preparation for GMR2 and swappable surfaces. Also maintain references to all surfaces in the command stream until the command stream has been fired in order to avoid racing with surface destruction taking place after validation but before submission. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Jakob Bornecrantz <jakob@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Michel Dänzer authored
Signed-off-by: Michel Dänzer <daenzer@vmware.com> Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Michel Dänzer authored
Signed-off-by: Michel Dänzer <daenzer@vmware.com> Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Michel Dänzer authored
Signed-off-by: Michel Dänzer <daenzer@vmware.com> Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Michel Dänzer authored
Signed-off-by: Michel Dänzer <daenzer@vmware.com> Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Thomas Hellstrom authored
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
- 31 Aug, 2011 2 commits
-
-
Marek Olšák authored
The new DRM_RADEON_GEM_WAIT ioctl combines GEM_WAIT_IDLE and GEM_BUSY (there is a NO_WAIT flag to get the latter) with USAGE_READ and USAGE_WRITE flags to take advantage of the new ttm_bo_wait changes. Also bump the DRM version. Signed-off-by: Marek Olšák <maraeo@gmail.com> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Marek Olšák authored
Sometimes we want to know whether a buffer is busy and wait for it (bo_wait). However, sometimes it would be more useful to be able to query whether a buffer is busy and being either read or written, and wait until it's stopped being either read or written. The point of this is to be able to avoid unnecessary waiting, e.g. if a GPU has written something to a buffer and is now reading that buffer, and a CPU wants to map that buffer for read, it needs to only wait for the last write. If there were no write, there wouldn't be any waiting needed. This, or course, requires user space drivers to send read/write flags with each relocation (like we have read/write domains in radeon, so we can actually use those for something useful now). Now how this patch works: The read/write flags should passed to ttm_validate_buffer. TTM maintains separate sync objects of the last read and write for each buffer, in addition to the sync object of the last use of a buffer. ttm_bo_wait then operates with one the sync objects. Signed-off-by: Marek Olšák <maraeo@gmail.com> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
- 30 Aug, 2011 4 commits
-
-
Rob Clark authored
Signed-off-by: Rob Clark <rob@ti.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Rob Clark authored
Signed-off-by: Rob Clark <rob@ti.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Alex Deucher authored
Fixes resume on Compaq Presario V5245EU. Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=41642Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@kernel.org Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Jerome Glisse authored
For some reason SPI block is in broken state after module unloading. This lead to broken rendering after reloading module. Fix this by reseting SPI block in CP resume function Signed-off-by: Jerome Glisse <jglisse@redhat.com Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@kernel.org Signed-off-by: Dave Airlie <airlied@redhat.com>
-
- 29 Aug, 2011 2 commits
-
-
Joonyoung Shim authored
It is left out the code to decrease the number of connector and encoder to the cleanup functions. Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Linus Torvalds authored
-
- 28 Aug, 2011 1 commit
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pmLinus Torvalds authored
* 'pm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: ARM: mach-shmobile: sh7372 LCDC1 suspend fix V2 (incremental) OMAP: omap_device: only override _noirq methods, not normal suspend/resume PM / Runtime: Correct documentation of pm_runtime_irq_safe() ARM: mach-shmobile: sh7372 LCDC1 suspend fix sh-sci / PM: Use power.irq_safe PM: Use spinlock instead of mutex in clock management functions
-
- 27 Aug, 2011 2 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6Linus Torvalds authored
* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6: firewire: sbp2: fix panic after rmmod with slow targets
-
Magnus Damm authored
This patch updates the recently submitted "Associate the HDMI clock together with LCDC1 on sh7372" to V2 with the following change: - Use lcdc1_device on AP4EVB to build properly. Signed-off-by: Magnus Damm <damm@opensource.se> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
-
- 26 Aug, 2011 12 commits
-
-
NeilBrown authored
The nfsservctl system call is now gone, so we should remove all linkage for it. Signed-off-by: NeilBrown <neilb@suse.de> Signed-off-by: J. Bruce Fields <bfields@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-
git://git.kernel.org/pub/scm/linux/kernel/git/keithp/linux-2.6Linus Torvalds authored
* 'drm-intel-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/keithp/linux-2.6: drm/i915: Fix wrong initializer for "locked" variable in assert_panel_unlocked i915: do not setup intel_backlight twice
-
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6Linus Torvalds authored
* 'usb-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6: (30 commits) USB: ftdi_sio: add Calao reference board support USB option driver K3765/K4505 avoid CDC_DATA interface USB: option: add YUGA device id to driver usb: s5p-ehci: fix a NULL pointer deference USB: EHCI: Do not rely on PORT_SUSPEND to stop USB resuming in ehci_bus_resume(). USB option driver add PID of Huawei Vodafone K4605 USB option driver add PID of Huawei Vodafone K3806 xhci: Handle zero-length isochronous packets. USB: Avoid NULL pointer deref in usb_hcd_alloc_bandwidth. usb: musb: gadget: fix error path usb: gadget: f_phonet: unlock in error case usb: musb: blackfin: include prefetch head file usb: musb: tusb6010: fix compilation usb: gadget: renesas_usbhs: fix DMA build by including dma-mapping.h usb: musb: cppi: fix build errors due to DBG and missing musb variable usb: musb: ux500: replace missing DBG with dev_dbg usb: musb: ux500: set dma config for both src and dst usb: musb: fix oops on musb_gadget_pullup usb: host: ehci-omap: fix .remove and failure handling path of .probe(v1) usb: gadget: hid: don't STALL when processing a HID Descriptor request ...
-
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty-2.6Linus Torvalds authored
* 'tty-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty-2.6: omap-serial: Allow IXON and IXOFF to be disabled. TTY: serial, document ignoring of uart->ops->startup error TTY: pty, fix pty counting 8250: Fix race condition in serial8250_backup_timeout(). serial/8250_pci: delete duplicate data definition 8250_pci: add support for Rosewill RC-305 4x serial port card tty: Add "spi:" prefix for spi modalias atmel_serial: fix atmel_default_console_device serial: 8250_pnp: add Intermec CV60 touchscreen device drivers/serial/ucc_uart.c: Fix compiler warning pch_uart: Set PCIe bus number using probe parameter serial: samsung: Fix build error
-
Linus Torvalds authored
Merge branch 'driver-core-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core-2.6 * 'driver-core-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core-2.6: drivers:misc: ti-st: fix unexpected UART close drivers:misc: ti-st: free skb on firmware download drivers:misc: ti-st: wait for completion at fail drivers:misc: ti-st: reinit completion before send drivers:misc: ti-st: fail-safe on wrong pkt type drivers:misc: ti-st: reinit completion on ver read drivers:misc:ti-st: platform hooks for chip states drivers:misc: ti-st: avoid a misleading dbg msg base/devres.c: quiet sparse noise about context imbalance pti: add missing CONFIG_PCI dependency drivers/base/devtmpfs.c: correct annotation of `setup_done' driver core: fix kernel-doc warning in platform.c firmware: fix google/gsmi.c build warning
-
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging-2.6Linus Torvalds authored
* 'staging-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging-2.6: staging: tidspbridge: fix compilation on dsp clock functions staging: octeon-ethernet: Add missing #includes. Staging: zcache: signedness bug in tmem_get() staging: zcache: fix crash on high memory swap staging: brcm80211: SPARC build error fix staging: brcm80211: fix compile error on non-x86 archs since 3.0 kernel
-
git://git390.marist.edu/pub/scm/linux-2.6Linus Torvalds authored
* 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6: [S390] memory hotplug: only unassign assigned increments [S390] Change default action from reipl to stop for on_restart [S390] arch/s390/kernel/ipl.c: correct error detection check [S390] drivers/s390/block/dasd_ioctl.c: add missing kfree [S390] nss,initrd: kernel image and initrd must be in different segments
-
Kirill A. Shutemov authored
According to the SFI specification irq number 0xFF means device has no interrupt or interrupt attached via GPIO. Currently, we don't handle this special case and set irq field in *_board_info structs to 255. It leads to confusion in some drivers. Accelerometer driver tries to register interrupt 255, fails and prints "Cannot get IRQ" to dmesg. Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-
git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6Linus Torvalds authored
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6: (32 commits) ALSA: hda: Conexant: Allow different output types to share DAC ASoC: Correct element count for WM8996 sidetone HPF ASoC: Tegra: wm8903 machine driver: Drop Ventana support ASoC: Add samsung maintainer ASoC: Add Springbank I/O card to Speyside Kconfig ALSA: hda/conexant - Enable ADC-switching for auto-mic mode, too ALSA: hda - Fix double-headphone/speaker paths for Cxt auto-parser ALSA: hda - Update jack-sense info even when no automute is set ALSA: hda - Fix output-path initialization for Realtek auto-parser sound/soc/fsl/mpc8610_hpcd.c: add missing of_node_put sound/soc/fsl/p1022_ds.c: add missing of_node_put sound/soc/ep93xx/ep93xx-i2s.c: add missing kfree sound/soc/kirkwood/kirkwood-i2s.c: add missing kfree ASoC: soc-core: use GFP_KERNEL flag for kmalloc in snd_soc_cnew sound/soc/fsl/fsl_dma.c: add missing of_node_put ASoC: Clear completions from late WM8996 FLL lock IRQs ASoC: Clear any outstanding WM8962 FLL lock completions before waiting ASoC: Ensure we only run Speyside WM8962 bias level callbacks once ASoC: Fix configuration of WM8996 input enables ASoC: WM8996 record paths need AIFCLK ...
-
Takashi Iwai authored
-
Liu Gang-B34182 authored
This bug causes the IECSR register clear failure. In this case, the RETE (retry error threshold exceeded) interrupt will be generated and cannot be cleared. So the related ISR may be called persistently. The RETE bit in IECSR is cleared by writing a 1 to it. Signed-off-by: Liu Gang <Gang.Liu@freescale.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Kumar Gala <galak@kernel.crashing.org> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-
MyungJoo Ham authored
The previous rtc-s3c had two issues related with its IRQ. 1. Users cannot open rtc multiple times because an open operation calls request_irq on the same IRQ. (e.g., two user processes wants to open and read RTC time from rtc-s3c at the same time) 2. If alarm is set and no one has the rtc opened with filesystem (either the alarm is set by kernel/boot-loader or user set an alarm and closed rtc dev file), the pending bit is not cleared and no further interrupt is invoked. When the alarm is used by the system itself such as a resume from suspend-to-RAM or other Low-power modes/idle, this is a critical issue. This patch mitigates these issues by calling request_irq at probe and free_irq at remove. Signed-off-by: MyungJoo Ham <myungjoo.ham@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Acked-by: Kukjin Kim <kgene.kim@samsung.com> Cc: Alessandro Zummo <a.zummo@towertech.it> Cc: Changhwan Youn <chaos.youn@samsung.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-