AuxXxilium/linux_dsm_epyc7002
1
0
Fork 0
mirror of https://github.com/AuxXxilium/linux_dsm_epyc7002.git synced 2025-02-10 12:05:21 +07:00
Code Issues Actions Packages Projects Releases Wiki Activity
09cf7c9a12
linux_dsm_epyc7002/drivers/gpu/drm
History
Chris Wilson 09cf7c9a12 drm/i915: Insert a flush between batches if the breadcrumb was dropped 
If we drop the breadcrumb request after a batch due to a signal for
example we aim to fix it up at the next opportunity. In this case we
emit a second batchbuffer with no waits upon the first and so no
opportunity to insert the missing request, so we need to emit the
missing flush for coherency. (Note that that invalidating the render
cache is the same as flushing it, so there should have been no
observable corruption.)

Note that beside simply adding the missing flush, avoiding potential
render corruption, this will also fix at least parts of the problem
introduced by some funny interaction of these two commits:

commit de2b998552
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Jul 4 22:52:50 2012 +0200

    drm/i915: don't return a spurious -EIO from intel_ring_begin

which allowed intel_ring_begin to return -ERESTARTSYS and

commit cc889e0f6c
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Jun 13 20:45:19 2012 +0200

    drm/i915: disable flushing_list/gpu_write_list

which essentially disabled the flushing list.

The issue happens when we submit a batch & emit it, but get
interrupted (thanks to the first patch) while trying to emit the
flush. On the next batch we still assume that the full gpu domain
handling is in effect and hence compute the invalidate&flushing
domains. But thanks to the 2nd patch we totally ignore these and only
invalidate all gpu domains, presuming that any required flushes have
been issued already.  Which is wrong and eventually results in us
updating the new write_domain values with the computed
pending_write_domain values, which leaves an object with write_domain
== 0 on the gpu_write_list.

As soon as we try to unbind that object, things blow up.

Fix this by emitting the missing flush according to the new
ring->gpu_caches_dirty flag.

Note that this does _not_ fix all the current cases where we end up
with an object on the flushing_list that can't be flushed.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=52040
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Add bug explanation to commit message.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2012-07-20 12:21:40 +02:00
..
ast drm: Make the .mode_fixup() operations mode argument a const pointer 2012-07-19 21:52:38 -04:00
cirrus drm: Make the .mode_fixup() operations mode argument a const pointer 2012-07-19 21:52:38 -04:00
exynos drm: Make the .mode_fixup() operations mode argument a const pointer 2012-07-19 21:52:38 -04:00
gma500 drm: kill reclaim_buffers callback 2012-07-19 22:50:28 -04:00
i2c drm: Make the .mode_fixup() operations mode argument a const pointer 2012-07-19 21:52:38 -04:00
i810 drm: kill dma queue support 2012-07-19 22:50:55 -04:00
i915 drm/i915: Insert a flush between batches if the breadcrumb was dropped 2012-07-20 12:21:40 +02:00
mga drm: kill reclaim_buffers callback 2012-07-19 22:50:28 -04:00
mgag200 drm/mgag200: fix null pointer dereference 2012-07-19 22:05:25 -04:00
nouveau drm: kill reclaim_buffers callback 2012-07-19 22:50:28 -04:00
r128 drm: kill reclaim_buffers callback 2012-07-19 22:50:28 -04:00
radeon drm: kill reclaim_buffers callback 2012-07-19 22:50:28 -04:00
savage drm/savage: clean up reclaim_buffers 2012-07-19 22:50:16 -04:00
sis drm/sis: fixup sis_mm ioctl structs 2012-07-19 22:51:58 -04:00
tdfx drm: kill reclaim_buffers callback 2012-07-19 22:50:28 -04:00
ttm drm/ttm: Fix buffer object metadata accounting regression v2 2012-06-12 15:57:47 +01:00
udl drm/udl: port over blanking code from udlfb. 2012-07-19 22:30:41 -04:00
via drm/via: clean up reclaim_buffers 2012-07-19 22:48:28 -04:00
vmwgfx drm: kill reclaim_buffers_locked 2012-07-19 22:49:58 -04:00
ati_pcigart.c
drm_agpsupport.c
drm_auth.c
drm_buffer.c
drm_bufs.c drm: kill dma queue support 2012-07-19 22:50:55 -04:00
drm_cache.c
drm_context.c
drm_crtc_helper.c drm: Don't initialize local ret variable when not needed 2012-05-22 10:32:58 +01:00
drm_crtc.c drm: Use stdint types for consistency 2012-05-29 11:07:09 +01:00
drm_debugfs.c drm: kill dma queue support 2012-07-19 22:50:55 -04:00
drm_dma.c drm: kill dma queue support 2012-07-19 22:50:55 -04:00
drm_dp_i2c_helper.c
drm_drv.c drm: kill dma queue support 2012-07-19 22:50:55 -04:00
drm_edid_load.c drm: Don't initialize local ret variable when not needed 2012-05-22 10:32:58 +01:00
drm_edid_modes.h
drm_edid.c drm/edid: don't return stack garbage from supports_rb 2012-06-21 09:34:13 +01:00
drm_encoder_slave.c
drm_fb_helper.c drm/fb helper: don't call drm_crtc_helper_set_config 2012-07-19 21:24:32 -04:00
drm_fops.c drm: unconditionally clean up dma buffers of closing clients 2012-07-19 22:51:04 -04:00
drm_gem.c drm: Add colouring to the range allocator 2012-07-16 05:59:37 +10:00
drm_global.c
drm_hashtab.c
drm_info.c drm: kill dma queue support 2012-07-19 22:50:55 -04:00
drm_ioc32.c
drm_ioctl.c drm/prime: expose capability flags for userspace. 2012-05-18 11:12:16 +01:00
drm_irq.c drm: Disallow DRM_IOCTL_MODESET_CTL for KMS drivers 2012-07-18 06:05:48 +10:00
drm_lock.c drm: ditch strange DRIVER_DMA_QUEUE only error bail-out 2012-07-19 22:50:47 -04:00
drm_memory.c
drm_mm.c drm: Add colouring to the range allocator 2012-07-16 05:59:37 +10:00
drm_modes.c
drm_pci.c drm/pci: add support for getting the supported link bw. 2012-07-19 22:29:25 -04:00
drm_platform.c
drm_prime.c drm/prime: add exported buffers to current fprivs imported buffer list (v2) 2012-05-23 10:46:03 +01:00
drm_proc.c drm: kill dma queue support 2012-07-19 22:50:55 -04:00
drm_scatter.c
drm_stub.c drm: Don't initialize local ret variable when not needed 2012-05-22 10:32:58 +01:00
drm_sysfs.c drm: fail gracefully when proc isn't setup. 2012-07-16 05:57:03 +10:00
drm_trace_points.c
drm_trace.h
drm_usb.c
drm_vm.c
Kconfig drm/kms: driver for virtual cirrus under qemu 2012-05-17 11:02:24 +01:00
Makefile drm/kms: driver for virtual cirrus under qemu 2012-05-17 11:02:24 +01:00
README.drm

README.drm 

************************************************************
* For the very latest on DRI development, please see:      *
*     http://dri.freedesktop.org/                          *
************************************************************

The Direct Rendering Manager (drm) is a device-independent kernel-level
device driver that provides support for the XFree86 Direct Rendering
Infrastructure (DRI).

The DRM supports the Direct Rendering Infrastructure (DRI) in four major
ways:

    1. The DRM provides synchronized access to the graphics hardware via
       the use of an optimized two-tiered lock.

    2. The DRM enforces the DRI security policy for access to the graphics
       hardware by only allowing authenticated X11 clients access to
       restricted regions of memory.

    3. The DRM provides a generic DMA engine, complete with multiple
       queues and the ability to detect the need for an OpenGL context
       switch.

    4. The DRM is extensible via the use of small device-specific modules
       that rely extensively on the API exported by the DRM module.


Documentation on the DRI is available from:
    http://dri.freedesktop.org/wiki/Documentation
    http://sourceforge.net/project/showfiles.php?group_id=387
    http://dri.sourceforge.net/doc/

For specific information about kernel-level support, see:

    The Direct Rendering Manager, Kernel Support for the Direct Rendering
    Infrastructure
    http://dri.sourceforge.net/doc/drm_low_level.html

    Hardware Locking for the Direct Rendering Infrastructure
    http://dri.sourceforge.net/doc/hardware_locking_low_level.html

    A Security Analysis of the Direct Rendering Infrastructure
    http://dri.sourceforge.net/doc/security_low_level.html