dri-devel May 2011

dri-devel@lists.freedesktop.org

118 participants
495 discussions

[PULL] drm-intel-next
by Keith Packard 03 Jun '11

03 Jun '11

Here's most of the patches I'm hoping to land in 2.6.40: * Ivybridge support (Gen7) * Forcewake fixes for Sandybridge (And ivybridge). * Temporary FB for load detect (hoping for some cleanups here) Still pending: * More modesetting cleanups (as always) * Disabling FBC on Ironlake to enable RC6 instead This sequence also assigns me as the the drm/i915 maintainer. The following changes since commit 2fb4e61d9471867677c97bf11dba8f1e9dfa7f7c: drm/i915/lvds: Only act on lid notify when the device is on (2011-05-09 09:13:22 -0700) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/keithp/linux-2.6.git drm-intel-next Ben Widawsky (10): drm/i915: fix ilk rc6 teardown locking drm/1915: ringbuffer wait for idle function drm/i915: fix rc6 initialization on Ironlake drm/i915: debugfs for context information drm/i915: proper use of forcewake drm/i915: reference counted forcewake drm/i915: forcewake struct mutex locking fixes drm/i915: move gen6 rps handling to workqueue drm/i915: debugfs interface for forcewake reference count drm/i915: forcewake debugfs fix Chris Wilson (12): drm/i915: Move the irq wait queue initialisation into the ring init drm/i915: Simplify return value from intel_get_load_detect_pipe drm/i915: Propagate failure to set mode for load-detect pipe drm/i915: Don't store temporary load-detect variables in the generic encoder drm/i915: Remove unused supported_crtc from intel_load_detect_pipe drm/i915: Pass the saved adjusted_mode when adding to the load-detect crtc drm/i915: Remove dead code from intel_get_load_detect_pipe() drm/i915: Remove dead code from intel_release_load_detect_pipe() drm/i915: Attach a fb to the load-detect pipe drm/i915: Rename agp_type to cache_level drm/i915: Do not clflush snooped objects drm/i915: Disable all outputs early, before KMS takeover Eric Anholt (12): drm/i915: Split the crtc_mode_set function along HAS_PCH_SPLIT() lines. drm/i915: Move the vblank pre/post modeset to the common crtc_mode_set. drm/i915: Remove the PCH paths from the pre-Ironlake crtc_mode_set(). drm/i915: Drop the eDP paths from the pre-Ironlake crtc_mode_set. drm/i915: Drop the remaining bit of Ironlake code from i9xx_crtc_mode_set(). drm/i915: Drop non-HAS_PCH_SPLIT() code from ironlake_crtc_mode_set(). drm/i915: Drop remaining pre-Ironlake code from ironlake_crtc_mode_set(). drm/i915: Clean up leftover DPLL and LVDS register choice from pch split. drm/i915: Fold the DPLL limit defines into the structs that use them. drm/i915: Use existing function instead of open-coding fence reg clear. drm/i915: Add support for fence registers on Ivybridge. drm/i915: Update the location of the ringbuffers' HWS_PGA registers for IVB. Jesse Barnes (20): drm/i915: use i915_enable_rc6 on SNB too drm/i915: make FDI training a display function drm/i915: split irq handling into per-chipset functions drm/i915: split enable/disable vblank code into chipset specific functions drm/i915: add IS_GEN7 macro to cover Ivy Bridge and later drm/i915: add IS_IVYBRIDGE macro for checks drm/i915: Ivy Bridge has split display and pipe control drm/i915: add swizzle/tiling support for Ivy Bridge drm/i915: manual FDI training for Ivy Bridge drm/i915: treat Ivy Bridge watermarks like Sandy Bridge drm/i915: interrupt & vblank support for Ivy Bridge drm/i915: page flip support for Ivy Bridge drm/i915: ring support for Ivy Bridge agp/intel: add Ivy Bridge support drm/i915: add PantherPoint PCH ID drm/i915: add Ivy Bridge PCI IDs and driver feature structs drm/i915: set IBX pch type explicitly drm/i915: split clock gating init into per-chipset functions drm/i915: add Ivybridge clock gating init function drm/i915: split PCH clock gating init Keith Packard (1): MAINTAINERS: Switch maintainer for drm/i915 to Keith Packard MAINTAINERS | 4 +- drivers/char/agp/intel-agp.c | 3 + drivers/char/agp/intel-agp.h | 8 + drivers/char/agp/intel-gtt.c | 10 + drivers/gpu/drm/i915/i915_debugfs.c | 128 ++- drivers/gpu/drm/i915/i915_dma.c | 60 +- drivers/gpu/drm/i915/i915_drv.c | 61 +- drivers/gpu/drm/i915/i915_drv.h | 111 +- drivers/gpu/drm/i915/i915_gem.c | 36 +- drivers/gpu/drm/i915/i915_gem_gtt.c | 35 +- drivers/gpu/drm/i915/i915_gem_tiling.c | 2 +- drivers/gpu/drm/i915/i915_irq.c | 310 ++++- drivers/gpu/drm/i915/i915_reg.h | 35 +- drivers/gpu/drm/i915/i915_suspend.c | 3 +- drivers/gpu/drm/i915/intel_crt.c | 24 +- drivers/gpu/drm/i915/intel_display.c | 2370 ++++++++++++++++++------------- drivers/gpu/drm/i915/intel_drv.h | 19 +- drivers/gpu/drm/i915/intel_ringbuffer.c | 38 +- drivers/gpu/drm/i915/intel_ringbuffer.h | 35 +- drivers/gpu/drm/i915/intel_tv.c | 13 +- 20 files changed, 2124 insertions(+), 1181 deletions(-) -- keith.packard(a)intel.com

1 4

[Bug 37785] New: [r600g Evergreen] GPU lockup on Blender 2.57 when moving objects
by bugzilla-daemon＠freedesktop.org 03 Jun '11

03 Jun '11

https://bugs.freedesktop.org/show_bug.cgi?id=37785 Summary: [r600g Evergreen] GPU lockup on Blender 2.57 when moving objects Product: DRI Version: XOrg CVS Platform: x86-64 (AMD64) OS/Version: Linux (All) Status: NEW Severity: normal Priority: medium Component: DRM/Radeon AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: ahlaht(a)gmail.com Created an attachment (id=47353) --> (https://bugs.freedesktop.org/attachment.cgi?id=47353) dmesg with errors GPU lockup (waiting for 0x00004E51 last fence id 0x00004E50) Using the latest git: Linux 2.6.39, drm 61be940, mesa 29ceeeb, xorg 1.10.2, ati 6.14.2 Hardware: HP laptop, Core i5-450M + Radeon HD 5650 REDWOOD (68C1) Steps to reproduce: 1. Start Blender 2.57 normally: $ blender-2.5 --factory-startup 2. On the default scene with the cube press 'g' to move the cube 3. Use mouse to continuously move the cube on screen on random directions 4. After about 5 seconds Blender freezes but mouse cursor keeps moving 5. A GPU lockup occurs and the provided errors appear on dmesg The bug is very easy to reproduce on my system. I can keep using the system normally after the bug, but sometimes I have to reset the video mode by switching to text console and then back to Xorg. Also there is an alternative way to trigger the bug: Instead of moving the cube keep resizing any of the side panels continuously. The problem disappears if I gpu-switcheroo to i965 so I assume this is something r600g related. Also I was not able to reproduce it on r600c. -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 2

3D support for Displaylink devices
by PRASANNA KUMAR 03 Jun '11

03 Jun '11

USB graphics devices from displaylink does not have 3D hardware. To get 3D effects (compiz, GNOME 3, KWin, OpenGL apps etc) with these device in Linux the native (primary) GPU can be used to provide hardware acceleration. All the graphics operation is done using the native (primary) GPU and the end result is taken and send to the displaylink device. Can this be achieved? If so is it possible to implement a generic framework so that any device (USB, thunderbolt or any new technology) can use this just by implementing device specific (compression and) data transport? I am not sure this is the correct mailing list. Thanks, Prasanna Kumar

5 5

Fw: [PATCH] drm: i915: correct return status in intel_hdmi_mode_valid()
by Nicolas Kaiser 02 Jun '11

02 Jun '11

Forwarding to Keith Packard. Sorry, initially sent to the previous maintainer instead. Start weitergeleitete Nachricht: Datum: Fri, 20 May 2011 18:58:53 +0200 Von: Nicolas Kaiser <nikai(a)nikai.net> An: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: David Airlie <airlied(a)linux.ie>, dri-devel(a)lists.freedesktop.org, linux-kernel(a)vger.kernel.org Betreff: [PATCH] drm: i915: correct return status in intel_hdmi_mode_valid() Signed-off-by: Nicolas Kaiser <nikai(a)nikai.net> --- Untested: just looks to me like the intention might be to return MODE_CLOCK_LOW here, is that correct? drivers/gpu/drm/i915/intel_hdmi.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index f289b86..655bbd9 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -194,7 +194,7 @@ static int intel_hdmi_mode_valid(struct drm_connector *connector, if (mode->clock > 165000) return MODE_CLOCK_HIGH; if (mode->clock < 20000) - return MODE_CLOCK_HIGH; + return MODE_CLOCK_LOW; if (mode->flags & DRM_MODE_FLAG_DBLSCAN) return MODE_NO_DBLESCAN; -- 1.7.5.rc3

2 1

[PATCH] drm/radeon: fix oops in ttm reserve when pageflipping (v2)
by Dave Airlie 31 May '11

31 May '11

From: Dave Airlie <airlied(a)redhat.com> We need to take a reference to this object, pinning doesn't take a reference so if userspace deletes the object it can disappear even if pinned. v2: fix error paths to unreference properly also. should fix: https://bugzilla.kernel.org/show_bug.cgi?id=32402 and https://bugzilla.redhat.com/show_bug.cgi?id=680651 Signed-off-by: Dave Airlie <airlied(a)redhat.com> --- drivers/gpu/drm/radeon/radeon_display.c | 13 ++++++++----- 1 files changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index ae247ee..292f73f 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -264,6 +264,8 @@ static void radeon_unpin_work_func(struct work_struct *__work) radeon_bo_unreserve(work->old_rbo); } else DRM_ERROR("failed to reserve buffer after flip\n"); + + drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base); kfree(work); } @@ -371,6 +373,8 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc, new_radeon_fb = to_radeon_framebuffer(fb); /* schedule unpin of the old buffer */ obj = old_radeon_fb->obj; + /* take a reference to the old object */ + drm_gem_object_reference(obj); rbo = gem_to_radeon_bo(obj); work->old_rbo = rbo; INIT_WORK(&work->work, radeon_unpin_work_func); @@ -378,12 +382,9 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc, /* We borrow the event spin lock for protecting unpin_work */ spin_lock_irqsave(&dev->event_lock, flags); if (radeon_crtc->unpin_work) { - spin_unlock_irqrestore(&dev->event_lock, flags); - kfree(work); - radeon_fence_unref(&fence); - DRM_DEBUG_DRIVER("flip queue: crtc already busy\n"); - return -EBUSY; + r = -EBUSY; + goto unlock_free; } radeon_crtc->unpin_work = work; radeon_crtc->deferred_flip_completion = 0; @@ -497,6 +498,8 @@ pflip_cleanup1: pflip_cleanup: spin_lock_irqsave(&dev->event_lock, flags); radeon_crtc->unpin_work = NULL; +unlock_free: + drm_gem_object_unreference_unlocked(old_radeon_fb->obj); spin_unlock_irqrestore(&dev->event_lock, flags); radeon_fence_unref(&fence); kfree(work); -- 1.7.4.4

2 1

[PATCH 1/4] drm/radeon/kms: add family ids for llano APUs
by Alex Deucher 31 May '11

31 May '11

Signed-off-by: Alex Deucher <alexdeucher(a)gmail.com> --- drivers/gpu/drm/radeon/radeon_device.c | 2 ++ drivers/gpu/drm/radeon/radeon_family.h | 2 ++ 2 files changed, 4 insertions(+), 0 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 5b61364..d77ede3 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -82,6 +82,8 @@ static const char radeon_family_name[][16] = { "CYPRESS", "HEMLOCK", "PALM", + "SUMO", + "SUMO2", "BARTS", "TURKS", "CAICOS", diff --git a/drivers/gpu/drm/radeon/radeon_family.h b/drivers/gpu/drm/radeon/radeon_family.h index 6f1d9e5..ec2f1ea 100644 --- a/drivers/gpu/drm/radeon/radeon_family.h +++ b/drivers/gpu/drm/radeon/radeon_family.h @@ -81,6 +81,8 @@ enum radeon_family { CHIP_CYPRESS, CHIP_HEMLOCK, CHIP_PALM, + CHIP_SUMO, + CHIP_SUMO2, CHIP_BARTS, CHIP_TURKS, CHIP_CAICOS, -- 1.7.1.1

1 3

[PATCH] drm/i915: add missing intel_enable_plane() call to i9xx_crtc_mode_set()
by Melchior FRANZ 31 May '11

31 May '11

From: Melchior FRANZ <mfranz(a)aon.at> Fix garbled up virtual linux console on Acer TM 5735Z-452G32Mnss (colored stripes, unreadable text) by adding the intel_enable_plane() call also to i9xx_crtc_mode_set(), which didn't inherit it from intel_crtc_mode_set() like its twin ironlake_crtc_mode_set(). Signed-off-by: Melchior FRANZ <mfranz(a)aon.at> --- The bug was originally introduced with 49183b2818 and fixed with 982b2035d9, but re-introduced by a recent drm/i915 branch merge. diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index f553ddf..5d20579 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -4735,6 +4735,8 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc, I915_WRITE(DSPCNTR(plane), dspcntr); POSTING_READ(DSPCNTR(plane)); + if (!HAS_PCH_SPLIT(dev)) + intel_enable_plane(dev_priv, plane, pipe); ret = intel_pipe_set_base(crtc, x, y, old_fb);

1 0

Re: [RFC PATCH 10/35] drivers/gpu changes for SMBIOS and System Firmware
by Chris Wilson 31 May '11

31 May '11

On Tue, 31 May 2011 11:51:47 -0400, Prarit Bhargava <prarit(a)redhat.com> wrote: > drivers/gpu changes for SMBIOS and System Firmware Acked-by: Chris Wilson <chris(a)chris-wilson.co.uk> -- Chris Wilson, Intel Open Source Technology Centre

1 0

[Bug 37781] New: Major redraw/flicker issues on rv350 after exiting fullscreen app with compiz running
by bugzilla-daemon＠freedesktop.org 31 May '11

31 May '11

https://bugs.freedesktop.org/show_bug.cgi?id=37781 Summary: Major redraw/flicker issues on rv350 after exiting fullscreen app with compiz running Product: Mesa Version: git Platform: Other OS/Version: All Status: NEW Severity: normal Priority: medium Component: Drivers/Gallium/r300 AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: oreaus(a)gmail.com With default packages in ubuntu natty 11.04, there is a problem where running and exiting a fullscreen application with compiz running causes major redraw/refresh flickering issues. They go away after restarting compiz, but start again reliably after exiting any fullscreen GL app. Here are the bisect results: 2a95568f64a6641a49a2d4855272e9be2ac2db6d is the first bad commit commit 2a95568f64a6641a49a2d4855272e9be2ac2db6d Author: Kenneth Graunke <kenneth(a)whitecape.org> Date: Wed May 11 07:54:57 2011 -0700 i965: Avoid register coalescing away MATH workarounds on Ivybridge. The MATH instruction cannot handle source modifiers, even on Gen7. So, apply this workaround for Sandybridge on Ivybridge as well. Signed-off-by: Kenneth Graunke <kenneth(a)whitecape.org> Reviewed-by: Eric Anholt <eric(a)anholt.net> :040000 040000 9cb73cbd09661d6ca745d42fa4aac07963d439b2 2b1804644fe873d2b054645c8e8d5a435d3cbe75 M src As a side note, the problem is actually worse in mesa current master, where it does not require triggering, but instead is flickering right after X/compiz start the first time. I plan to bisect between the problem commit here and master next to find out what causes it to get worse. -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 6

[Bug 21582] Build failure if NOUVEAU && !HWMON
by bugzilla-daemon＠bugzilla.kernel.org 30 May '11

30 May '11

https://bugzilla.kernel.org/show_bug.cgi?id=21582 Guenter Roeck <linux(a)roeck-us.net> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |linux(a)roeck-us.net --- Comment #1 from Guenter Roeck <linux(a)roeck-us.net> 2011-05-30 20:27:23 --- Fixed with commit b54262f3c828ee17e27632d0d60255281c02e1a5. -- Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are watching the assignee of the bug.

1 0

← Newer
1
...
27
28
29
30
31
32
33
...
50
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel May 2011