dri-devel April 2015

dri-devel@lists.freedesktop.org

175 participants
1044 discussions

Re: [Mesa-dev] st_TexSubImage: unaligned memcpy performance
by Vasilis Liaskovitis 09 Apr '15

09 Apr '15

Hi, On Wed, Apr 8, 2015 at 1:24 PM, Daniel Stone <daniel(a)fooishbar.org> wrote: > Hi, > > On 8 April 2015 at 10:57, Vasilis Liaskovitis <vliaskov(a)gmail.com> wrote: > > I have an issue where st_TexSubImage causes very high CPU load in > > __memcpy_sse2_unaligned (Mesa 10.1.3, Xorg 1.15.1, radeon driver, HD > 7870). > > > > Any obvious causes / tips for this? e.g. align textures or use different > > format/type? I 've tried using GL_BGRA/GL_UNSIGNED_BYTE and > > GL_BGRA/GL_UNSIGNED_INT_8_8_8_8_REV > > > > __memcpy_sse2_unaligned () at > > ../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S:85 > > 85 ../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S: No such file > or > > directory. > > (gdb) bt > > #0 __memcpy_sse2_unaligned () at > > ../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S:85 > > #1 0x00007fffb572f154 in memcpy (__len=7680, __src=<optimized out>, > > __dest=0x7fff5835f800) at /usr/include/x86_64-linux-gnu/bits/string3.h:51 > > #2 st_TexSubImage (ctx=0x1b91420, dims=<optimized out>, > texImage=0x1f81710, > > xoffset=0, yoffset=0, zoffset=0, width=1920, height=1080, depth=1, > > format=32993, type=5121, pixels=0xdacf90, unpack=0x1bad590) > > at ../../../../src/mesa/state_tracker/st_cb_texture.c:752 > > Your source (0xdacf90) is only aligned to a 16-byte boundary, not 32. > This will cause issues particularly on ARM, where natural alignment is > required (i.e. 32-byte load/stores must be on 32-byte boundaries). By > contrast, the destination is already aligned to a 128-byte boundary. > So fixing the caller, rather than Mesa, should take care of the > problem. > thanks for the reply and the observation. I aligned source on 32-byte boundary (or even 128-byte boundary) but there was no difference. By the way, I am only using x86_64, not ARM. I believe intel sse2 only requires 16-byte boundary alignment, but perhaps i am missing something. Is this code path in st_TexSubImage using PBOs? I guess it depends on driver (radeon in my case) implementation? Related: pboUnpack http://www.songho.ca/opengl/files/pboUnpack.zip gives: Transfer Rate: 236.5 MB/s. (59.1 FPS) Does this sounds reasonably ok for uploading with PBO? Same bottleneck __memcpy_sse2_unaligned is observed. sample perf report output: 28,20% pboUnpack libc-2.19.so [.] __memcpy_sse2_unaligned 16,63% pboUnpack pboUnpack [.] 0x0000000000006542 6,96% pboUnpack [kernel.kallsyms] [k] clear_page_c_e 2,52% pboUnpack [drm] [k] drm_mm_insert_node_in_range_generic 2,10% pboUnpack [kernel.kallsyms] [k] get_page_from_freelist backtrace: __memcpy_sse2_unaligned () at ../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S:86 86 ../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S: No such file or directory. (gdb) bt #0 __memcpy_sse2_unaligned () at ../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S:86 #1 0x00007ffff2bddbbd in memcpy (__len=4194304, __src=<optimized out>, __dest=<optimized out>) at /usr/include/x86_64-linux-gnu/bits/string3.h:51 #2 memcpy_texture (dimensions=dimensions@entry=2, dstFormat=dstFormat@entry=MESA_FORMAT_B8G8R8A8_UNORM, dstRowStride=dstRowStride@entry=4096, dstSlices=dstSlices@entry=0x7fffffffd6e8, srcWidth=srcWidth@entry=1024, srcHeight=srcHeight@entry=1024, srcDepth=srcDepth@entry=1, srcFormat=srcFormat@entry=32993, srcType=srcType@entry=5121, srcAddr=srcAddr@entry=0x7fffeeecd000, srcPacking=srcPacking@entry=0x7ffff7f69180, ctx=<optimized out>) at ../../../../src/mesa/main/texstore.c:949 #3 0x00007ffff2be353d in _mesa_texstore_memcpy (srcPacking=0x7ffff7f69180, srcAddr=<optimized out>, srcType=5121, srcFormat=32993, srcDepth=<optimized out>, srcHeight=<optimized out>, srcWidth=<optimized out>, dstSlices=<optimized out>, dstRowStride=<optimized out>, dstFormat=MESA_FORMAT_B8G8R8A8_UNORM, baseInternalFormat=6408, dims=<optimized out>, ctx=0x7ffff7f4d010) at ../../../../src/mesa/main/texstore.c:3938 #4 _mesa_texstore (ctx=0x7ffff7f4d010, dims=2, baseInternalFormat=6408, dstFormat=MESA_FORMAT_B8G8R8A8_UNORM, dstRowStride=4096, dstSlices=0x7fffffffd6e8, srcWidth=1024, srcHeight=1024, srcDepth=1, srcFormat=32993, srcType=5121, srcAddr=0x7fffeeecd000, srcPacking=0x7ffff7f69180) at ../../../../src/mesa/main/texstore.c:3958 #5 0x00007ffff2be3812 in store_texsubimage (ctx=ctx@entry=0x7ffff7f4d010, texImage=texImage@entry=0x7c8690, xoffset=xoffset@entry=0, yoffset=yoffset@entry=0, zoffset=zoffset@entry=0, width=1024, height=1024, depth=1, format=32993, type=5121, pixels=0x0, packing=0x7ffff7f69180, caller=0x7ffff2d609c7 "glTexSubImage") at ../../../../src/mesa/main/texstore.c:4107 #6 0x00007ffff2be3aa5 in _mesa_store_texsubimage (ctx=ctx@entry=0x7ffff7f4d010, dims=<optimized out>, texImage=texImage@entry=0x7c8690, xoffset=xoffset@entry=0, yoffset=yoffset@entry=0, zoffset=zoffset@entry=0, width=<optimized out>, width@entry=1024, height=<optimized out>, height@entry=1024, depth=<optimized out>, depth@entry=1, format=<optimized out>, format@entry=32993, type=<optimized out>, type@entry=5121, pixels=<optimized out>, pixels@entry=0x0, packing=<optimized out>, packing@entry=0x7ffff7f69180) at ../../../../src/mesa/main/texstore.c:4171 #7 0x00007ffff2c3acaa in st_TexSubImage (ctx=0x7ffff7f4d010, dims=<optimized out>, texImage=0x7c8690, xoffset=0, yoffset=0, zoffset=0, width=1024, height=1024, depth=1, format=32993, type=5121, pixels=0x0, unpack=0x7ffff7f69180) at ../../../../src/mesa/state_tracker/st_cb_texture.c:787 #8 0x00007ffff2bce83d in texsubimage (ctx=0x7ffff7f4d010, dims=dims@entry=2, target=3553, level=0, xoffset=0, yoffset=0, zoffset=zoffset@entry=0, width=1024, height=1024, depth=depth@entry=1, format=format@entry=32993, type=type@entry=5121, pixels=pixels@entry=0x0) at ../../../../src/mesa/main/teximage.c:3445 #9 0x00007ffff2bd259c in _mesa_TexSubImage2D (target=<optimized out>, level=<optimized out>, xoffset=<optimized out>, yoffset=<optimized out>, width=<optimized out>, height=<optimized out>, format=32993, type=5121, pixels=0x0) at ../../../../src/mesa/main/teximage.c:3483 pixels pointer in st_texSubImage is 0x0 here, maybe because it's an internal pbo to texture transfer? srcAddr in memcpy_texture() is 0x7fffeeecd000 which looks sufficiently aligned, but maybe this is not the correct pointer to look at. could there also be a CPU stall/sync issue when mapping a pbo buffer? Similar pbounpack/memcpy performance discussed a bit here recently with no conclusion: http://people.freedesktop.org/~cbrill/dri-log/?channel=dri-devel&date=2015-… thanks, - Vasilis > > Cheers, > Daniel >

1 0

[PATCH] drm/tegra: Remove unused .mode_set and .mode_set_base CRTC helpers
by Laurent Pinchart 09 Apr '15

09 Apr '15

The two CRTC helper operations are called only for non-atomic mode setting, by either the drm_crtc_helper_set_config() helper or the drm_helper_resume_force_mode() helper. As the driver has switched to atomic mode setting and neither of those helpers is used, the operations are not used anymore. Remove them. Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas(a)ideasonboard.com> --- drivers/gpu/drm/tegra/dc.c | 2 -- 1 file changed, 2 deletions(-) Hi Thierry, I stumbled on this while trying to understand the atomic mode setting code paths. Could you please test the patch ? diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c index 3aaa84ae2681..4476d6a35a0f 100644 --- a/drivers/gpu/drm/tegra/dc.c +++ b/drivers/gpu/drm/tegra/dc.c @@ -1327,9 +1327,7 @@ static void tegra_crtc_atomic_flush(struct drm_crtc *crtc) static const struct drm_crtc_helper_funcs tegra_crtc_helper_funcs = { .disable = tegra_crtc_disable, .mode_fixup = tegra_crtc_mode_fixup, - .mode_set = drm_helper_crtc_mode_set, .mode_set_nofb = tegra_crtc_mode_set_nofb, - .mode_set_base = drm_helper_crtc_mode_set_base, .prepare = tegra_crtc_prepare, .commit = tegra_crtc_commit, .atomic_check = tegra_crtc_atomic_check, -- Regards, Laurent Pinchart

3 2

[PATCH] drm/nouveau/bios: 0 means success for nvbios_extend
by Jan Vesely 09 Apr '15

09 Apr '15

Bugzilla:https://bugs.freedesktop.org/show_bug.cgi?id=89047 CC: David Airlie <airlied(a)linux.ie> CC: Ben Skeggs <bskeggs(a)redhat.com> CC: dri-devel(a)lists.freedesktop.org CC: linux-kernel(a)vger.kernel.org Signed-off-by: Jan Vesely <jano.vesely(a)gmail.com> --- It's needed for 3.19 too drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadow.c | 2 +- drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowacpi.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadow.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadow.c index 8c2b7cb..f566ac8 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadow.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadow.c @@ -44,7 +44,7 @@ shadow_fetch(struct nvkm_bios *bios, u32 upto) const u32 limit = (upto + 3) & ~3; const u32 start = bios->size; void *data = mthd->data; - if (nvbios_extend(bios, limit) > 0) { + if (nvbios_extend(bios, limit) >= 0) { u32 read = mthd->func->read(data, start, limit - start, bios); bios->size = start + read; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowacpi.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowacpi.c index 1fbd93b..f9d0eb5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowacpi.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowacpi.c @@ -52,7 +52,7 @@ acpi_read_fast(void *data, u32 offset, u32 length, struct nvkm_bios *bios) u32 start = offset & ~0x00000fff; u32 fetch = limit - start; - if (nvbios_extend(bios, limit) > 0) { + if (nvbios_extend(bios, limit) >= 0) { int ret = nouveau_acpi_get_bios_chunk(bios->data, start, fetch); if (ret == fetch) return fetch; @@ -73,7 +73,7 @@ acpi_read_slow(void *data, u32 offset, u32 length, struct nvkm_bios *bios) u32 start = offset & ~0xfff; u32 fetch = 0; - if (nvbios_extend(bios, limit) > 0) { + if (nvbios_extend(bios, limit) >= 0) { while (start + fetch < limit) { int ret = nouveau_acpi_get_bios_chunk(bios->data, start + fetch, -- 2.0.5

1 0

[git pull] drm fixes
by Dave Airlie 08 Apr '15

08 Apr '15

Hi Linus, Final drm fixes, one core locking imbalance regression, and a bunch of i915 baytrail s/r fixes. going to be away for a few days, but should still have email at least. Dave. The following changes since commit f22e6e847115abc3a0e2ad7bb18d243d42275af1: Linux 4.0-rc7 (2015-04-06 15:39:45 -0700) are available in the git repository at: git://people.freedesktop.org/~airlied/linux drm-fixes for you to fetch changes up to f4274e23fb16721449d973ed607505f5dbdcd2df: Merge tag 'drm-intel-fixes-2015-04-08' of git://anongit.freedesktop.org/drm-intel into drm-fixes (2015-04-09 06:59:50 +1000) ---------------------------------------------------------------- Dave Airlie (1): Merge tag 'drm-intel-fixes-2015-04-08' of git://anongit.freedesktop.org/drm-intel into drm-fixes Deepak S (1): drm/i915/chv: Remove Wait for a previous gfx force-off Jesse Barnes (2): drm/i915/vlv: save/restore the power context base reg drm/i915/vlv: remove wait for previous GFX clk disable request Tommi Rantala (1): drm: fix drm_mode_getconnector() locking imbalance regression drivers/gpu/drm/drm_crtc.c | 4 +++- drivers/gpu/drm/i915/i915_drv.c | 14 ++------------ drivers/gpu/drm/i915/i915_drv.h | 1 + 3 files changed, 6 insertions(+), 13 deletions(-)

1 0

[PATCH v3] ASoC: max98090: add shutdown callback for max98090
by Caesar Wang 08 Apr '15

08 Apr '15

To fix pop noise when shutdown,the pop noise during shutdown is the pmic cutoff power of codec without any notice. Signed-off-by: jay.xu <xjq(a)rock-chips.com> Signed-off-by: zhengxing <zhengxing(a)rock-chips.com> Signed-off-by: Caesar Wang <wxt(a)rock-chips.com> --- Changes in v3: - modify the shutdown function before remove(..) - fix the `Serien-cc` Changes in v2: - remove the dev_info(..) - fix the comment style - add the max98090_i2c_shutdown() in remove() fuction sound/soc/codecs/max98090.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/sound/soc/codecs/max98090.c b/sound/soc/codecs/max98090.c index b112b1c..3e33ef2 100644 --- a/sound/soc/codecs/max98090.c +++ b/sound/soc/codecs/max98090.c @@ -2605,8 +2605,24 @@ err_enable: return ret; } +static void max98090_i2c_shutdown(struct i2c_client *i2c) +{ + struct max98090_priv *max98090 = dev_get_drvdata(&i2c->dev); + + /* + * Enable volume smoothing, disable zero cross. This will cause + * a quick 40ms ramp to mute on shutdown. + */ + regmap_write(max98090->regmap, + M98090_REG_LEVEL_CONTROL, M98090_VSENN_MASK); + regmap_write(max98090->regmap, + M98090_REG_DEVICE_SHUTDOWN, 0x00); + msleep(40); +} + static int max98090_i2c_remove(struct i2c_client *client) { + max98090_i2c_shutdown(client); snd_soc_unregister_codec(&client->dev); return 0; } @@ -2696,6 +2712,7 @@ static struct i2c_driver max98090_i2c_driver = { .acpi_match_table = ACPI_PTR(max98090_acpi_match), }, .probe = max98090_i2c_probe, + .shutdown = max98090_i2c_shutdown, .remove = max98090_i2c_remove, .id_table = max98090_i2c_id, }; -- 1.9.1

2 1

[PATCH 1/1] drm/exynos: Fix FIMD buffer size calculation
by Daniel Stone 08 Apr '15

08 Apr '15

Commit adacb228d72b ("drm: Exynos: Respect framebuffer pitch for FIMD/Mixer") fixed the buffer size calculation by using the FB pitch value but later commit 26b9c2813ede1 ("drm/exynos: remove struct *_win_data abstraction on planes") added a regression so fix the buffer size calculation again. Tested on Chromebook Snow / Peach Pit. Fixes: 26b9c2813ede1 ("drm/exynos: remove struct *_win_data abstraction on planes") Signed-off-by: Daniel Stone <daniels(a)collabora.com> Tested-by: Javier Martinez Canillas <javier.martinez(a)collabora.co.uk> --- drivers/gpu/drm/exynos/exynos_drm_fimd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c index 7964b278eefb..90111b87c963 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c +++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c @@ -662,7 +662,7 @@ static void fimd_win_commit(struct exynos_drm_crtc *crtc, unsigned int win) writel(val, ctx->regs + VIDWx_BUF_START(win, 0)); /* buffer end address */ - size = plane->pitch * plane->crtc_height * (plane->bpp >> 3); + size = plane->pitch * plane->crtc_height; val = (unsigned long)(dma_addr + size); writel(val, ctx->regs + VIDWx_BUF_END(win, 0)); @@ -672,7 +672,7 @@ static void fimd_win_commit(struct exynos_drm_crtc *crtc, unsigned int win) plane->crtc_width, plane->crtc_height); /* buffer size */ - buf_offsize = (plane->fb_width - plane->crtc_width) * (plane->bpp >> 3); + buf_offsize = plane->pitch - (plane->crtc_width * (plane->bpp >> 3)); line_size = plane->crtc_width * (plane->bpp >> 3); val = VIDW_BUF_SIZE_OFFSET(buf_offsize) | VIDW_BUF_SIZE_PAGEWIDTH(line_size) | -- 2.1.4

2 1

[PATCH v2] ASoC: max98090: add shutdown callback for max98090
by Caesar Wang 08 Apr '15

08 Apr '15

To fix pop noise when shutdown,the pop noise during shutdown is the pmic cutoff power of codec without any notice. Signed-off-by: jay.xu <xjq(a)rock-chips.com> Signed-off-by: zhengxing <zhengxing(a)rock-chips.com> Signed-off-by: Caesar Wang <wxt(a)rock-chips.com> Serien-cc: linux-kernel(a)vger.kernel.org Serien-cc: devicetree(a)vger.kernel.org Serien-cc: dianders(a)chromium.org --- Changes in v2: - remove the dev_info(..) - fix the comment style - add the max98090_i2c_shutdown() in remove() fuction sound/soc/codecs/max98090.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/sound/soc/codecs/max98090.c b/sound/soc/codecs/max98090.c index b112b1c..e59013a 100644 --- a/sound/soc/codecs/max98090.c +++ b/sound/soc/codecs/max98090.c @@ -2607,10 +2607,26 @@ err_enable: static int max98090_i2c_remove(struct i2c_client *client) { + max98090_i2c_shutdown(client); snd_soc_unregister_codec(&client->dev); return 0; } +static void max98090_i2c_shutdown(struct i2c_client *i2c) +{ + struct max98090_priv *max98090 = dev_get_drvdata(&i2c->dev); + + /* + * Enable volume smoothing, disable zero cross. This will cause + * a quick 40ms ramp to mute on shutdown. + */ + regmap_write(max98090->regmap, + M98090_REG_LEVEL_CONTROL, M98090_VSENN_MASK); + regmap_write(max98090->regmap, + M98090_REG_DEVICE_SHUTDOWN, 0x00); + msleep(40); +} + #ifdef CONFIG_PM static int max98090_runtime_resume(struct device *dev) { @@ -2697,6 +2713,7 @@ static struct i2c_driver max98090_i2c_driver = { }, .probe = max98090_i2c_probe, .remove = max98090_i2c_remove, + .shutdown = max98090_i2c_shutdown, .id_table = max98090_i2c_id, }; -- 1.9.1

2 2

[PATCH 0/1] ASoC: max98090: add shutdown callback for max98090
by Caesar Wang 08 Apr '15

08 Apr '15

Tested on veyron devices,play music then shutdown device,here carefully to speaker or headphone. Caesar Wang (1): ASoC: max98090: add shutdown callback for max98090 sound/soc/codecs/max98090.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) -- 1.9.1

3 6

[PATCH 1/4] drm/radeon: add extra check in radeon_ttm_tt_unpin_userptr
by Christian König 08 Apr '15

08 Apr '15

From: Christian König <christian.koenig(a)amd.com> We somehow try to free the SG table twice. Bugs: https://bugs.freedesktop.org/show_bug.cgi?id=89734 Signed-off-by: Christian König <christian.koenig(a)amd.com> Cc: <stable(a)vger.kernel.org> --- drivers/gpu/drm/radeon/radeon_ttm.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index d02aa1d..b292aca 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -598,6 +598,10 @@ static void radeon_ttm_tt_unpin_userptr(struct ttm_tt *ttm) enum dma_data_direction direction = write ? DMA_BIDIRECTIONAL : DMA_TO_DEVICE; + /* double check that we don't free the table twice */ + if (!ttm->sg->sgl) + return; + /* free the sg table and pages again */ dma_unmap_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction); -- 1.9.1

2 5

[Bug 89734] GL_AMD_pinned_memory extension causing a kernel hardlock
by bugzilla-daemon＠freedesktop.org 08 Apr '15

08 Apr '15

https://bugs.freedesktop.org/show_bug.cgi?id=89734 Bug ID: 89734 Summary: GL_AMD_pinned_memory extension causing a kernel hardlock Product: Mesa Version: git Hardware: x86-64 (AMD64) OS: Linux (All) Status: NEW Severity: normal Priority: medium Component: Drivers/Gallium/radeonsi Assignee: dri-devel(a)lists.freedesktop.org Reporter: poub365-bugzilla(a)yahoo.com QA Contact: dri-devel(a)lists.freedesktop.org Kubuntu 15.04, kernel 4.0rc3, amd hd7850 with recent (from ppa oibaf) open source drivers. Dolphin emulator runs fine except that I have a complete freeze when I exit. I have to reboot the computer (although it seems to be still running as I can remotely log with ssh). I tried under kde, then under icewm, to see if it was the compositor's fault. No change. Windowed or fullscreen: same freeze. I tried to stop Dolphin with the interface buttons: same freeze. kubuntu 15.04, i7 4770k non overclocked, amd hd7850 on open source drivers. dmesg log is here: http://pastebin.ca/2962286 I reported the bug to the Dolphin team. Told me that "Basically it's AMD's GL_AMD_pinned_memory extension causing a kernel hardlock." and that I have to report here. -- You are receiving this mail because: You are the assignee for the bug.

1 18

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel April 2015