ttm_bo_mmap ttm_mem_reg_is_pci
Thomas there is 2 more bugs:
In ttm_bo_mmap if the bo = NULL after ttm_bo_vm_lookup_rb
then
if (unlikely(bo == NULL)) {
printk(KERN_ERR Could not find buffer object to map.\n);
goto out_unref;
}
Should be
if (unlikely(bo == NULL)) {
printk(KERN_ERR Could not find buffer object to map.\n);
return -EINVAL;
}
Otherwise you unref a null bo and bad things happen.
In ttm_mem_reg_is_pci i think it should return false if
flags TTM_MEMTYPE_FLAG_NEEDS_IOREMAP is not set otherwise
kmap on a gart (unmappable to cpu) will think it needs
to use ioremap which fails.
Cheers,
Jerome Glisse gli...@freedesktop.org
--
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
--
___
Dri-devel mailing list
Dri-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dri-devel
[Bug 12899] Crash in i915.ko: i915_driver_irq_handler
http://bugzilla.kernel.org/show_bug.cgi?id=12899 --- Comment #2 from helge.bahm...@secunet.com 2009-03-22 06:38 --- Are you sure this is sane? I mean, i915_gem_cleanup_hws sets dev_priv-hw_status_page=NULL without any locks held (that I am aware of, at least), and the irq handler is accessing the value without any locks either. The check for dev_priv-hw_status_page!=0 added into the irq handler IMHO just shortens the race window to the few instructions between performing the test and the subsequent READ_BREADCRUMB and thus hides the problem better. Is there any guard between i915_gem_cleanup_hws and the irq handler racing which I am missing so far? -- Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are the assignee for the bug, or are watching the assignee. -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: ttm_bo_mmap ttm_mem_reg_is_pci
Jerome Glisse wrote:
Thomas there is 2 more bugs:
In ttm_bo_mmap if the bo = NULL after ttm_bo_vm_lookup_rb
then
if (unlikely(bo == NULL)) {
printk(KERN_ERR Could not find buffer object to map.\n);
goto out_unref;
}
Should be
if (unlikely(bo == NULL)) {
printk(KERN_ERR Could not find buffer object to map.\n);
return -EINVAL;
}
Otherwise you unref a null bo and bad things happen.
In ttm_mem_reg_is_pci i think it should return false if
flags TTM_MEMTYPE_FLAG_NEEDS_IOREMAP is not set otherwise
kmap on a gart (unmappable to cpu) will think it needs
to use ioremap which fails.
Cheers,
Jerome Glisse gli...@freedesktop.org
Thanks, Jerome.
I'll take a look at these.
/Thomas
--
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
--
___
Dri-devel mailing list
Dri-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: ttm_bo_mmap ttm_mem_reg_is_pci
Jerome Glisse wrote: In ttm_mem_reg_is_pci i think it should return false if flags TTM_MEMTYPE_FLAG_NEEDS_IOREMAP is not set otherwise kmap on a gart (unmappable to cpu) will think it needs to use ioremap which fails. Jerome, in this case, how is the memory type backing the gart set up? There is a flag TTM_MEMTYPE_FLAG_CMA (CMA for Can't Map Aperture) that should map the underlying pages instead of the aperture in most relevant cases. /Thomas Cheers, Jerome Glisse gli...@freedesktop.org -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
newttm: drm_mm_put_block NULL
Hi Thomas, another bugs is that call to drm_mm_put_block should be followed by setting the ptr to NULL (at least if i understand this mm stuff) ie : drm_mm_put_block(bo-vm_node); should be followed by: bo-vm_node = NULL; I am sorry to not have a tree where you can grab fixes but right now i am in the middle of debugging radeon with newttm, but once i get it working i will push it (hopefully i should exhaust major bugs pretty soon). Cheers, Jerome Glisse gli...@freedesktop.org -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
newttm: ttm_buffer_object_transfer last take
Hi Thomas, There is another bug in ttm_buffer_object_transfer when we copy *fbo = *bo we should set the fbo-vm_node to NULL otherwise the vm_node will be free with the ghost object and bad things will happen for the still alive object. Cheers, Jerome Glisse gli...@freedesktop.org -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
[PATCH] drm/intel: only set TV mode when any property changed
If there's no real property change, don't need to set TV mode again.
Signed-off-by: Zhenyu Wang zhenyu.z.w...@intel.com
---
drivers/gpu/drm/i915/intel_tv.c | 27 ---
1 files changed, 20 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c
index ceca947..cd9d74c 100644
--- a/drivers/gpu/drm/i915/intel_tv.c
+++ b/drivers/gpu/drm/i915/intel_tv.c
@@ -1571,32 +1571,45 @@ intel_tv_set_property(struct drm_connector *connector,
struct drm_property *prop
struct intel_output *intel_output = to_intel_output(connector);
struct intel_tv_priv *tv_priv = intel_output-dev_priv;
int ret = 0;
+ bool changed = false;
ret = drm_connector_property_set_value(connector, property, val);
if (ret 0)
goto out;
- if (property == dev-mode_config.tv_left_margin_property)
+ if (property == dev-mode_config.tv_left_margin_property
+ tv_priv-margin[TV_MARGIN_LEFT] != val) {
tv_priv-margin[TV_MARGIN_LEFT] = val;
- else if (property == dev-mode_config.tv_right_margin_property)
+ changed = true;
+ } else if (property == dev-mode_config.tv_right_margin_property
+ tv_priv-margin[TV_MARGIN_RIGHT] != val) {
tv_priv-margin[TV_MARGIN_RIGHT] = val;
- else if (property == dev-mode_config.tv_top_margin_property)
+ changed = true;
+ } else if (property == dev-mode_config.tv_top_margin_property
+ tv_priv-margin[TV_MARGIN_TOP] != val) {
tv_priv-margin[TV_MARGIN_TOP] = val;
- else if (property == dev-mode_config.tv_bottom_margin_property)
+ changed = true;
+ } else if (property == dev-mode_config.tv_bottom_margin_property
+ tv_priv-margin[TV_MARGIN_BOTTOM] != val) {
tv_priv-margin[TV_MARGIN_BOTTOM] = val;
- else if (property == dev-mode_config.tv_mode_property) {
+ changed = true;
+ } else if (property == dev-mode_config.tv_mode_property) {
if (val = NUM_TV_MODES) {
ret = -EINVAL;
goto out;
}
+ if (!strcmp(tv_priv-tv_format, tv_modes[val].name))
+ goto out;
+
tv_priv-tv_format = tv_modes[val].name;
- intel_tv_mode_set(intel_output-enc, NULL, NULL);
+ changed = true;
} else {
ret = -EINVAL;
goto out;
}
- intel_tv_mode_set(intel_output-enc, NULL, NULL);
+ if (changed)
+ intel_tv_mode_set(intel_output-enc, NULL, NULL);
out:
return ret;
}
--
1.5.6.5
--
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
--
___
Dri-devel mailing list
Dri-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dri-devel
[Bug 11877] [i965 texture_srgb] texture_srgb glGetTexImage failed for internalFormat GL_SRGB_EXT
http://bugs.freedesktop.org/show_bug.cgi?id=11877 --- Comment #17 from Shuang He shuang...@intel.com 2009-03-22 20:27:53 PST --- this issue still exists with: Kernel_version: 2.6.29-rc7 Libdrm: (master)2e2e8575b1ed4703653a72ac2b60b75316c388d7 Mesa:(mesa_7_4_branch)a8528a2e8653b5237c1d1d66fe98c6e031d007f9 Xserver: (server-1.6-branch)60c161545af80eb78eb790a05bde79409dfdf16e Xf86_video_intel: (2.7)238c2c40afd9f8b61479b8640d53f20d52fd7ddf Kernel: (for-airlied)dc529a4fe1ae4667c819437a94185e8581e1e680 -- Configure bugmail: http://bugs.freedesktop.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug. -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
[Bug 13759] [i965 FBO GEM] render to 1D texture work incorrecly
http://bugs.freedesktop.org/show_bug.cgi?id=13759 --- Comment #5 from Shuang He shuang...@intel.com 2009-03-22 20:28:36 PST --- this issue still exists with: Kernel_version: 2.6.29-rc7 Libdrm: (master)2e2e8575b1ed4703653a72ac2b60b75316c388d7 Mesa:(mesa_7_4_branch)a8528a2e8653b5237c1d1d66fe98c6e031d007f9 Xserver: (server-1.6-branch)60c161545af80eb78eb790a05bde79409dfdf16e Xf86_video_intel: (2.7)238c2c40afd9f8b61479b8640d53f20d52fd7ddf Kernel: (for-airlied)dc529a4fe1ae4667c819437a94185e8581e1e680 -- Configure bugmail: http://bugs.freedesktop.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug. -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
[Bug 13801] [i965 FBO GEM] render to Mipmap texture through FBO is not working
http://bugs.freedesktop.org/show_bug.cgi?id=13801 --- Comment #3 from Shuang He shuang...@intel.com 2009-03-22 20:29:17 PST --- this issue still exists with: Kernel_version: 2.6.29-rc7 Libdrm: (master)2e2e8575b1ed4703653a72ac2b60b75316c388d7 Mesa:(mesa_7_4_branch)a8528a2e8653b5237c1d1d66fe98c6e031d007f9 Xserver: (server-1.6-branch)60c161545af80eb78eb790a05bde79409dfdf16e Xf86_video_intel: (2.7)238c2c40afd9f8b61479b8640d53f20d52fd7ddf Kernel: (for-airlied)dc529a4fe1ae4667c819437a94185e8581e1e680 -- Configure bugmail: http://bugs.freedesktop.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug. -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com -- ___ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
