Re: [PATCH v3 1/1] drm/amdgpu: export gfx config double offchip LDS buffers (v3)

Am 20.02.2017 um 10:03 schrieb Zhang, Jerry: Hi Christian, Read the mail again, maybe you'd like to add amdgpu_gfx_config to the amdgpu_gca_config, right? No, like you did it in the patch is already fine. I don't see much of a reason for two structures here. But you could rename

Re: [PATCH V2] drm/amdgpu: Fix module unload hang by KIQ IRQ set

will be invoked earlier than that of AMD_IP_BLOCK_TYPE_IH. So continue to use the variable freed by AMD_IP_BLOCK_TYPE_GFX will cause system hang. Changes in v2: -- Refine according to Xiangliang Yu's suggestions Signed-off-by: Trigger Huang <trigger.hu...@amd.com> Acked-by: Christian

Re: [PATCH] drm/ttm: fix use-after-free races in vm fault handling

m> Good catch! Patch is Reviewed-by: Christian König <christian.koe...@amd.com> -- This does not fix the random memory corruption I've been seeing. --- drivers/gpu/drm/ttm/ttm_bo_vm.c | 12 1 file changed, 12 insertions(+) diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/

Re: [PATCH] drm/amdgpu: trace amdgpu_job fence details

Am 23.02.2017 um 21:48 schrieb Andres Rodriguez: On 2017-02-23 02:46 PM, Andres Rodriguez wrote: On 2017-02-23 03:20 AM, Christian König wrote: Am 23.02.2017 um 00:47 schrieb Andres Rodriguez: This trace is intended to provide the required information to associate the completion

Re: [PATCH 1/2] drm/amdgpu: make trace format uniform csv name=value

Am 24.02.2017 um 19:20 schrieb Andres Rodriguez: Most of the traces have uniform format except for two of them. Having all the traces match makes it simple to run awk on the ftrace output. Signed-off-by: Andres Rodriguez <andre...@gmail.com> Reviewed-by: Christian König <chri

Re: [PATCH 2/2] drm/amdgpu: trace fence details in amdgpu_sched_run_job

m_ibs=%u", If you have time for another patch please drop the pinters here as well. Scheduler job, IBs and fence are all heavily reallocated (sometimes even with a slab allocator). So the pointers are completely meaningless. The adev pointer is superseded by the timeline name and context number

Re: [PATCH] drm/amdgpu: trace amdgpu_job fence details

Am 23.02.2017 um 00:47 schrieb Andres Rodriguez: This trace is intended to provide the required information to associate the completion of an amdgpu_job with its corresponding dma_fence_* tracepoints. Signed-off-by: Andres Rodriguez ---

Re: [PATCH] drm/amdgpu: fix double_offchip_lds_buf for gfx v6

Am 23.02.2017 um 04:07 schrieb Alex Deucher: On Wed, Feb 22, 2017 at 10:04 PM, Junwei Zhang <jerry.zh...@amd.com> wrote: Signed-off-by: Junwei Zhang <jerry.zh...@amd.com> Reviewed-by: Alex Deucher <alexander.deuc...@amd.com> Reviewed-by: Christian König <chr

Re: [PATCH] drm/amd/amdgpu: Fix flow control in uvd_v4_2_stop()

Am 13.02.2017 um 20:13 schrieb Tom St Denis: Break out of outer loop properly. Signed-off-by: Tom St Denis <tom.stde...@amd.com> Reported-by: Dan Carpenter <dan.carpen...@oracle.com> Reviewed-by: Christian König <christian.koe...@amd.com>. --- drivers/gpu/drm/amd/amd

Re: [PATCH] drm/amdgpu: Remove bo.priority compiler warning

Yeah, Nicolai was faster than you. Patch is already reviewed and should be committed. Christian. Am 14.02.2017 um 19:30 schrieb Felix Kuehling: I think Nicolai beat you to this with his patch "drm/ttm: make TTM_MAX_BO_PRIORITY unsigned". On 17-02-14 01:03 PM, Kent Russell wrote: Addresses

Re: How to catch wild writes to system memory, and IOMMU errors

Uff, well that's quite a problem you ran into here. IOMMU might not help here, cause when it would be the GPU we would have made a mapping once and then the page in question is never unmapped (IIRC). To confirm if it is really the GPU writing those bytes I would add a trace point to

Re: [PATCH] drm/ttm: make TTM_MAX_BO_PRIORITY unsigned

peof(y), \ ^ drivers/gpu/drm/amd/amdgpu/amdgpu_object.c:434:21: note: in expansion of macro ‘min’ bo->tbo.priority = min(bo->tbo.priority, TTM_MAX_BO_PRIORITY - 1); ^~~ Signed-off-by: Nicolai Hähnle <nicolai.haeh...@amd.com> Reviewed-by: Christian Kö

Re: PRT support for amdgpu v3

T mappings exist, I get a WARN_ON (which seems fairly straightforward) and occasionally also an RCU error warning -- see the attached dmesg logs. Regular application shutdown works fine, though. Cheers, Nicolai On 08.02.2017 16:04, Christian König wrote: Hi guys, ok I finally found time to wri

Re: [PATCH v2 2/2] drm/amdgpu: report the number of bytes moved at buffer creation

Am 13.02.2017 um 17:28 schrieb Nicolai Hähnle: On 09.02.2017 11:33, Samuel Pitoiset wrote: Like ttm_bo_validate(), ttm_bo_init() might need to move BO and the number of bytes moved by TTM should be reported. This can help the throttle buffer migration mechanism to make a better decision. Hmm,

Re: [PATCH 1/2] drm/amdgpu: fix PRT cleanup order in the VM

Am 13.02.2017 um 18:32 schrieb Nicolai Hähnle: On 13.02.2017 17:40, Nicolai Hähnle wrote: On 13.02.2017 14:23, Christian König wrote: From: Christian König <christian.koe...@amd.com> We need to unmap the PRTs first and then free our scheduler entity. Thanks for the quick fix! Both p

Re: [PATCH v2 1/2] drm/amdgpu: fix a potential deadlock in amdgpu_bo_create_restricted()

Am 13.02.2017 um 19:32 schrieb Samuel Pitoiset: On 02/13/2017 07:19 PM, Nicolai Hähnle wrote: On 13.02.2017 19:11, Samuel Pitoiset wrote: On 02/13/2017 07:04 PM, Nicolai Hähnle wrote: On 13.02.2017 18:49, Samuel Pitoiset wrote: On 02/13/2017 05:25 PM, Nicolai Hähnle wrote: On

Re: [PATCH] drm/amdgpu:use hw_init for sriov_gpu_reset

for sriov_gpu_reset no suspend invoked so after VF FLR by host, we just call hw_init to reinitialize IPs. Change-Id: If09cb42b09bee6acc84e6b239ef537ad5a3df41c Signed-off-by: Monk Liu <monk@amd.com> Reviewed-by: Alex Deucher <alexander.deuc...@amd.com> Reviewed-by: Christian König &l

Re: PRT support for amdgpu v3

Am 09.02.2017 um 11:11 schrieb Nicolai Hähnle: On 08.02.2017 16:04, Christian König wrote: Hi guys, ok I finally found time to write an unit test for this and hammered out the last few bugs. Seems to work fine on my Tonga now. Please note that this set is based on "fix race in GEM V

Re: [PATCH 1/2] drm/amdgpu: drop multiple bo_unreserve() calls in amdgpu_gem_op_ioctl()

NAK, we need to drop the reservation before calling copy_to_user(). Otherwise you run into a double locking error when copy_to_user() is trying to copy to the BO in question. Regards, Christian. Am 09.02.2017 um 23:28 schrieb Samuel Pitoiset: Move amdgpu_bo_unreserve() outside of the

Re: [PATCH] drm/amdgpu: add more cases to DCE11 possible crtc mask setup

el.org Reviewed-by: Christian König <christian.koe...@amd.com>. --- drivers/gpu/drm/amd/amdgpu/dce_v11_0.c | 6 ++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index 1cf1d9d..5b24e89 100644 --- a/drivers/gpu/dr

Re: [PATCH 2/2] drm/amdgpu: use amdgpu_gem_va_check() in amdgpu_gem_va_update_vm()

Am 09.02.2017 um 23:28 schrieb Samuel Pitoiset: This removes code duplication. Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> Reviewed-by: Christian König <christian.koe...@amd.com>. --- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 8 +--- 1 file changed, 1 ins

Re: [PATCH 1/2] drm/amdgpu: drop multiple bo_unreserve() calls in amdgpu_gem_op_ioctl()

Am 10.02.2017 um 11:11 schrieb Samuel Pitoiset: On 02/10/2017 03:55 AM, zhoucm1 wrote: On 2017年02月10日 06:28, Samuel Pitoiset wrote: Move amdgpu_bo_unreserve() outside of the switch. While we are at it, add a missing break in the default case. Signed-off-by: Samuel Pitoiset

Re: [PATCH 1/1] drm/amdgpu: export gfx capability by gpu info

Am 16.02.2017 um 10:24 schrieb Zhang, Jerry: -Original Message- From: Christian König [mailto:deathsim...@vodafone.de] Sent: Thursday, February 16, 2017 17:12 To: Zhang, Jerry; amd-gfx@lists.freedesktop.org Subject: Re: [PATCH 1/1] drm/amdgpu: export gfx capability by gpu info Am

Re: [PATCH 1/1] drm/amdgpu: export gfx capability by gpu info

Am 16.02.2017 um 03:53 schrieb Junwei Zhang: Change-Id: Ibf3e4dbb7deb83271adabc275c9b7a0e0652541a Signed-off-by: Junwei Zhang Complete NAK on this. The drm_amdgpu_capability structure is specific to the pro driver. The Vulkan stack on the other hand is supposed to work

Re: [PATCH v3 3/3] drm/amdgpu: simplify reservation handling during buffer creation

ttm_bo_unreserve(>tbo); You can just use amdgpu_bo_unreserve() here. With that fixed the whole set is Reviewed-by: Christian König <christian.koe...@amd.com>. Regards, Christian. *bo_ptr = bo; trace_amdgpu_bo_create(bo); __

Re: [PATCH 1/3] drm/amdgpu: minor PRT turnoff fix

Am 15.02.2017 um 18:28 schrieb Nicolai Hähnle: Hi Christian, On 15.02.2017 16:59, Christian König wrote: Nicolai could you give that set a try? It should fix your problems with PRT tear down on process crash. Yes, it fixes those issues for me, thanks! The first two patches have my R-b

Re: [PATCH 1/3] drm/amdgpu: minor PRT turnoff fix

Nicolai could you give that set a try? It should fix your problems with PRT tear down on process crash. Regards, Christian. Am 15.02.2017 um 15:57 schrieb Christian König: From: Christian König <christian.koe...@amd.com> When two VMs stop using PRT support at the same time we

Re: [PATCH] drm/amdgpu: change pte definitions to 64 bit

Am 15.02.2017 um 19:09 schrieb Alex Xie: Change-Id: Iaeeb253cc5c028c93c536ee6ea08213c7fa8f299 Signed-off-by: Alex Xie <alexbin@amd.com> Ah, good catch. I've missed that one. Reviewed-by: Christian König <christian.koe...@amd.com>. --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.

[PATCH 1/3] drm/amdgpu: minor PRT turnoff fix

From: Christian König <christian.koe...@amd.com> When two VMs stop using PRT support at the same time we might not disable it in the right order otherwise. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 +--- 1 fil

[PATCH 3/3] drm/amdgpu: fix PRT teardown on VM fini v2

From: Christian König <christian.koe...@amd.com> v2: new approach fixing this by registering a fence callback for all users of the VM on teardown Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgp

[PATCH 2/3] drm/amdgpu: add OOM fallback on PRT teardown

From: Christian König <christian.koe...@amd.com> Don't assume kmalloc will always succeed. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 28 ++-- 1 file changed, 22 insertions(+), 6 deletions(-)

[PATCH 1/2] drm/amdgpu: fix 64bit shift for KV/KB

From: Christian König <christian.koe...@amd.com> Fixes "access stolen VRAM directly on KV/KB" otherwise that won't work as expected with more than 4GB RAM. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c | 2 +- 1 fil

[PATCH 2/2] drm/amdgpu: fix 64bit shift for CZ

From: Christian König <christian.koe...@amd.com> Fixes "access stolen VRAM directly on CZ", otherwise that won't work as expected with more than 4GB RAM. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 2 +- 1 file chan

Re: [PATCH] Add missing CIK devices

Am 13.02.2017 um 13:46 schrieb Tom St Denis: Adds mullins, kabini, and hawaii ASICs to the library. Signed-off-by: Tom St Denis <tom.stde...@amd.com> Acked-by: Christian König <christian.koe...@amd.com>. --- src/lib/asic/CMakeLists.txt | 3 +++ src/lib/asic/hawaii.

[PATCH 2/2] drm/amdgpu: fix PTE defines

From: Christian König <christian.koe...@amd.com> Those should be 64bit, even on a 32bit system. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 14 +++--- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/dri

[PATCH 1/2] drm/amdgpu: fix PRT cleanup order in the VM

From: Christian König <christian.koe...@amd.com> We need to unmap the PRTs first and then free our scheduler entity. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)

Re: [PATCH] drm/amdgpu: refuse to reserve io mem for split VRAM buffers

Am 17.02.2017 um 00:21 schrieb Nicolai Hähnle: On 17.02.2017 00:02, Alex Deucher wrote: On Thu, Feb 16, 2017 at 5:55 PM, Nicolai Hähnle wrote: From: Nicolai Hähnle When the fast blit path fails while attempting to move a buffer from RAM to VRAM,

Re: [PATCH 1/1] drm/amdgpu: export gfx capability by gpu info

(so they stay zero cleared). Not sure what the UMD is expecting here. Regards, Christian. Regards, Jerry (Junwei Zhang) Linux Base Graphics SRDC Software Development _ -Original Message- From: Christian König [mailto:deathsim...@vodafone.de] Sent

Re: [PATCH v2 1/1] drm/amdgpu: export gfx config by gpu info (v2)

Am 17.02.2017 um 06:05 schrieb Junwei Zhang: v2: move the config struct to drm_amdgpu_info_device Signed-off-by: Junwei Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 7 +++ drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 1 + drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c

Re: [PATCH 1/2] drm/amdgpu: fix reboot failure issue for virtualization

Am 18.01.2017 um 06:00 schrieb Xiangliang Yu: Reboot process will call HW fini functions of IP blocks. For virt, need to send event three before hw fini and send event four after hw fini. Signed-off-by: Xiangliang Yu <xiangliang...@amd.com> Reviewed-by: Christian König <chri

Re: [PATCH 0/3] some cleanups and fixes

Reviewed-by: Christian König <christian.koe...@amd.com> for the series. Regards, Christian. Am 18.01.2017 um 03:40 schrieb Monk Liu: Monk Liu (3): drm/amdgpu:set cond_exec polling value to 1 in ring_init drm/amdgpu:Preamble is forbid to be ignored in SRIOV drm/amdgpu:insert

Re: 答复: [PATCH] drm/amdgpu:guarantee 128dws between vm flush and IB

rmance. BR Monk ---- *发件人:* Christian König <deathsim...@vodafone.de> *发送时间:* 2017年1月18日 17:28:15 *收件人:* Liu, Monk; amd-gfx@lists.freedesktop.org *主题:* Re: [PATCH] drm/amdgpu:guarantee 128dws between vm flush and IB Am 18.01.2017 u

Re: [PATCH 1/1] drm/amdgpu: fix unload driver issue for virtual display

Am 19.01.2017 um 03:03 schrieb Xiangliang Yu: Virtual display doesn't allocate amdgpu_encoder when initializing, so will get invaild pointer if try to free amdgpu_encoder when unloading driver. Signed-off-by: Xiangliang Yu <xiangliang...@amd.com> Acked-by: Christian König <chri

Re: 答复: 答复: [PATCH] drm/amdgpu:guarantee 128dws between vm flush and IB

*发件人:* amd-gfx <amd-gfx-boun...@lists.freedesktop.org> 代表 Christian König <deathsim...@vodafone.de> *发送时间:* 2017年1月18日 20:52:14 *收件人:* Liu, Monk; amd-gfx@lists.freedesktop.org *主题:* Re: 答复: [PATCH] drm/amdgpu:guarantee 128dws between vm f

Re: [PATCH] drm/amdgpu:guarantee 128dws between vm flush and IB(v3)

Am 18.01.2017 um 12:42 schrieb Monk Liu: previously we always insert 128nops behind vm_flush, which may lead to DAMframe size above 256 dw and automatially aligned to 512 dw. now we calculate how many DWs already inserted after vm_flush and make up for the reset to pad up to 128dws before

Re: [PATCH] drm/amdgpu/virt: increase mailbox timeout value

Acked-by: Christian König <christian.koe...@amd.com>. Am 19.01.2017 um 11:56 schrieb Ding, Pixel: Reviewed-by: Pixel Ding <pixel.d...@amd.com> — Sincerely Yours, Pixel On 19/01/2017, 5:46 PM, "amd-gfx on behalf of Xiangliang Yu" <amd-gfx-boun...@lists.

Re: [PATCH] drm/amdgpu:guarantee 128dws between vm flush and IB(v3)

Am 19.01.2017 um 14:51 schrieb Grazvydas Ignotas: On Thu, Jan 19, 2017 at 11:10 AM, Christian König <deathsim...@vodafone.de> wrote: Am 18.01.2017 um 12:42 schrieb Monk Liu: @@ -6743,6 +6741,15 @@ static void gfx_v8_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t

Re: 答复: [PATCH 2/3] drm/amdgpu:implement META-DATA write routines

: if ring_write_multiple is organized in a separate patch, doesn't it introduces an function that no client using it ?? fine by me although ... BR Monk *发件人:* Christian König <deathsim...@vodafone.de> *发送时间:* 2017年1月12日 20

Re: [PATCH 0/5] META-DATA init packages patches v2

Patch #1 is Acked-by: Christian König <christian.koe...@amd.com>. Patch #2 - #4 are Reviewed-by: Christian König <christian.koe...@amd.com>. Patch #5: As Alex noted you need to update the dw estimation for gfx_v8_ring_emit_cntxcntl() as well when you add this. With that fix

Re: [PATCH] drm/amdgpu:in cntx_ctrl we need insert meta-init for CE/DE(V2)

viewed-by: Christian König <christian.koe...@amd.com>. Regards, Christian. --- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 10 +- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index fdc2cb5..52

Re: [PATCH] drm/amdgpu: simplify allocation of scratch regs

-by: Nils Wallménius <nils.wallmen...@gmail.com> Reviewed-by: Christian König <christian.koe...@amd.com>. --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 3 +-- drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 21 +++-- drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c | 7 +--

Re: [PATCH] drm/amdgpu: check ring being ready before using

second time it is good practice to add "v2" to the subject and something like "v2: changed xyz" to the commit message. Signed-off-by: Ding Pixel <pd...@amd.com> Reviewed-by: Christian König <christian.koe...@amd.com>. --- drivers/gpu/drm/amd/amdgpu/amdgp

Re: [PATCH libdrm 1/3] amdgpu: verify the tested device

Am 20.01.2017 um 04:45 schrieb zhoucm1: some small nitpick... On 2017年01月20日 06:53, Alex Xie wrote: Verify the vender ID and driver name. Open all AMDGPU devices. Provide an option to open render node. Tested as root: PASS Tested as non-privileged user: All tests failed as expected

Re: [PATCH] drm/amdgpu:guarantee 128dws between vm flush and IB(v3)

, Monk Sent: Thursday, January 19, 2017 10:39 PM To: 'Christian König' <deathsim...@vodafone.de>; amd-gfx@lists.freedesktop.org Subject: RE: [PATCH] drm/amdgpu:guarantee 128dws between vm flush and IB(v3) Christian I just found one issue in your previous patch ( the one you remove inserting

Re: [PATCH drm/amdgpu] Use correct mask in dce_v8_0_afmt_setmode() and fix comment typos.

that fixed the patch is Reviewed-by: Christian König <christian.koe...@amd.com>. Regards, Christian. WREG32(mmAFMT_RAMP_CONTROL0 + offset, 0x00FF); WREG32(mmAFMT_RAMP_CONTROL1 + offset, 0x007F); WREG32(mmAFMT_RAMP_CONTROL2 + offset, 0x0001);

Re: [PATCH drm/amdgpu] Fix indentation in dce_v8_0_audio_write_sad_regs()

Am 22.08.2016 um 02:41 schrieb Edward O'Callaghan: On 08/22/2016 10:38 AM, Alexandre Demers wrote: Fixed indentation for readability. Signed-off-by: Alexandre Demers <alexandre.f.dem...@gmail.com> Reviewed-by: Edward O'Callaghan <funfunc...@folklore1984.net> Reviewed-by: Chr

Re: Random short freezes due to TTM buffer migrations

Am 16.08.2016 um 17:56 schrieb Marek Olšák: On Tue, Aug 16, 2016 at 5:27 PM, Christian König <deathsim...@vodafone.de> wrote: Hi Marek, I'm already working on this. My current approach is to use a custom BO manager for VRAM with TTM and so split allocations into chunks of 4MB. Lar

Re: Random short freezes due to TTM buffer migrations

exporting it, I think that should work. Regards, Felix On 16-08-17 02:58 AM, Christian König wrote: One question: Will it be possible to share these split BOs as dmabufs? In theory yes, in practice I'm not sure. DMA-bufs are designed around scatter gather tables, those fortunately support

Re: [PATCH v2 0/3] drm/amdgpu: update golden setting of VI

Well, usually they are just the recommended settings from the hardware team which try to fix or improve things by changing the default values for the registers. I work for AMD and sometimes it isn't clear to me either what those changes are good for. E.g. I would need to dig though a whole

Re: [PATCH v2 2/2] drm/amdgpu: expose AMDGPU_GEM_CREATE_VRAM_CLEARED to user space

Am 01.08.2016 um 15:59 schrieb Alex Deucher: On Mon, Aug 1, 2016 at 1:27 AM, Flora Cui <flora@amd.com> wrote: I'll push if no objection Reviewed-by: Alex Deucher <alexander.deuc...@amd.com> Sorry for the delay, been on vacation last week. Patch is Reviewed-by: Chr

Re: [PATCH 0/3] VCE3 soft reset working

Reviewed-by: Christian König <christian.koe...@amd.com> for the whole series. Regards, Christian. Am 03.08.2016 um 09:51 schrieb Chunming Zhou: VCE is always busy is since RB1_BUSY is always busy, now remove it. And enable VCE soft reset. Chunming Zhou (2): drm/amdgpu: fix harvest

Re: [PATCH 2/2] drm/amdgpu: trace fence details in amdgpu_sched_run_job

Am 25.02.2017 um 18:28 schrieb Andres Rodriguez: On Feb 25, 2017 4:40 AM, "Christian König" <deathsim...@vodafone.de <mailto:deathsim...@vodafone.de>> wrote: Am 24.02.2017 um 19:20 schrieb Andres Rodriguez: This information is intended to pro

Re: [PATCH 15/22] drm/amdgpu: add a mechanism to untie user ring ids from kernel ring ids

Am 02.03.2017 um 09:02 schrieb Andres Rodriguez: Add amdgpu_queue_mgr, a mechanism that allows disjointing usermode's ring ids from the kernel's ring ids. The queue manager maintains a per-file descriptor map of user ring ids to amdgpu_ring pointers. Once a map is created it is permanent (this

Re: Add support for high priority scheduling in amdgpu

Am 01.03.2017 um 18:24 schrieb Andres Rodriguez: On 2017-03-01 12:13 PM, Andres Rodriguez wrote: On 3/1/2017 6:42 AM, Christian König wrote: Patches #1-#14 are Acked-by: Christian König <christian.koe...@amd.com>. Patch #15: Not sure if that is a good idea or not, need to take a

Re: [PATCH] drm/ttm: fix use-after-free races in vm fault handling

Am 26.02.2017 um 22:35 schrieb Daniel Vetter: On Sun, Feb 19, 2017 at 10:32:43AM +0100, Christian König wrote: Am 18.02.2017 um 23:50 schrieb Nicolai Hähnle: From: Nicolai Hähnle <nicolai.haeh...@amd.com> The vm fault handler relies on the fact that the VMA owns a reference to the BO. H

Re: [PATCH 0/6] make ctx mgr global

Am 24.08.2016 um 11:33 schrieb zhoucm1: On 2016年08月18日 16:15, Christian König wrote: NAK to the whole approach. If we want to share dependencies in the form of fences between devices and especially processes we must use android fences and the sync file framework. Then if we want to share

Re: [PATCH] drm/amdgpu:fix DMAframe for GFX8

fresh new, and Assume DE is still in VMSWITCH-TO-P2 stage (P1/P2 means process1/process2), since this time vm page table of P2 yet not been binding to VMID3, if CE is already running in IB belongs to P2, why we never meet vm fault ? Thanks for your time! BR Monk -Original Mess

Re: [PATCH] drm/amdgpu:fix DMAframe for GFX8

| > DE CE > Still assume the ring buffer is fresh new, and Assume DE is still in VMSWITCH-TO-P2 stage (P1/P2 means process1/process2), since this time vm page table of P2 yet not been binding to VMID3, if CE is already running in IB bel

Re: [PATCH 0/6] make ctx mgr global

Am 24.08.2016 um 12:01 schrieb zhoucm1: On 2016年08月24日 17:39, Christian König wrote: Am 24.08.2016 um 11:33 schrieb zhoucm1: On 2016年08月18日 16:15, Christian König wrote: NAK to the whole approach. If we want to share dependencies in the form of fences between devices and especially

Re: [PATCH 0/6] make ctx mgr global

Yeah, I wouldn't mind a patch renaming ctx to fence_context in the job and rign structure. This way we make it clear what is used here. Christian. Am 25.08.2016 um 05:38 schrieb Liu, Monk: Sorry, I don't mean ctx-id, I'm talking about "job->ctx", which is a atomic_64t value, and it is

Re: [PATCH 2/2] drm/amdgpu:256dw is enough for one ib_schedule

Well userspace is allowed to send any number of IBs down to this if I remember correctly. So we clearly need something depending on the number of IBs or reject submissions with to many IBs earlier in the CS. Otherwise we clearly open up a possible problem where userspace can trigger a ring

Re: [PATCH 2/2] drm/amdgpu:256dw is enough for one ib_schedule

buffer is only 4kb size -Original Message- From: Christian König [mailto:deathsim...@vodafone.de] Sent: Thursday, August 25, 2016 4:12 PM To: Liu, Monk <monk@amd.com>; amd-gfx@lists.freedesktop.org Subject: Re: [PATCH 2/2] drm/amdgpu:256dw is enough for one ib_schedule Well use

Re: 答复: [PATCH 2/2] drm/amdgpu/vce3: add support for third vce ring

Reviewed-by: Christian König <christian.koe...@amd.com> as well. Do we have any requirement to support ring 2 and 3 in the near future or was this set just for cleanup? Regards, Christian. Am 25.08.2016 um 02:03 schrieb Qu, Jim: These series Reivewed-by: JimQu <jim...@amd.com>

Re: [PATCH] drm/amdgpu:implement CONTEXT_CONTROL (v3)

perly clean up this feature while maintaining backward and forward compatibility. Regards, Christian. BR Monk BR Monk -Original Message- From: Bas Nieuwenhuizen [mailto:b...@basnieuwenhuizen.nl] Sent: Friday, September 02, 2016 12:09 AM To: Liu, Monk <monk....@amd.com&

Re: [PATCH 2/2] drm/amdgpu: bind GTT on demand

Am 06.09.2016 um 11:53 schrieb zhoucm1: On 2016年09月06日 17:41, Christian König wrote: From: Christian König <christian.koe...@amd.com> We don't really need the GTT table any more most of the time. Why? I thought GTT bo is always needed to be bound when GPU is trying to access it, d

Re: [PATCH] drm/amdgpu:implement CONTEXT_CONTROL (v3)

1) Is that my patch must work with current MESA driver ? (even MESA doesn't change any bit of its logic) ? Yes and it must work with all old versions of Mesa. This is usual the tricky part to get right and most of my concern right now. 2) is that my patch can let kmd go to a new path (using

[PATCH 1/2] drm/amdgpu: fix GTT offset handling

From: Christian König <christian.koe...@amd.com> Otherwise we run into problems on 32bit systems with more than 4GB GART. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 10 +- 1 file changed, 5 insertions(+), 5 deleti

[PATCH 2/2] drm/amdgpu: bind GTT on demand

From: Christian König <christian.koe...@amd.com> We don't really need the GTT table any more most of the time. So bind it only on demand. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu.h| 1 + drivers/gpu/drm/amd/amdgpu

Re: [PATCH] drm/amd/amdgpu: Remove double lock from gfx v6

Denis <tom.stde...@amd.com> Reviewed-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c | 4 1 file changed, 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index a6b76170b9c5..

Re: [Updated PATCH v2] drm: modify drm_global_item_ref to avoid two times of writing ref->object

Am 07.09.2016 um 09:54 schrieb Huang Rui: On Wed, Sep 07, 2016 at 09:12:29AM +0200, Christian König wrote: Am 07.09.2016 um 07:24 schrieb Huang Rui: In previous drm_global_item_ref, there are two times of writing ref->object if item->refcount is 0. So this patch does a minor update

Re: [PATCH v3] drm: modify drm_global_item_ref to avoid two times of writing ref->object

Am 08.09.2016 um 09:35 schrieb Chris Wilson: On Thu, Sep 08, 2016 at 03:22:48PM +0800, Huang Rui wrote: On Thu, Sep 08, 2016 at 02:36:06PM +0800, Chris Wilson wrote: On Wed, Sep 07, 2016 at 10:07:57PM -0400, Huang Rui wrote: In previous drm_global_item_ref, there are two times of writing

Re: [PATCH v3] drm: modify drm_global_item_ref to avoid two times of writing ref->object

times of writing ref->object. It can make the code logic more clearly. Signed-off-by: Huang Rui <ray.hu...@amd.com> Reviewed-by: Christian König <christian.koe...@amd.com>. --- Changes from V2 -> V3: - Use duplicate mutex release to avoid "goto" in non-error patch.

Re: [PATCH] amd/powerplay: use kmemdup instead of kmalloc + memcpy

Am 04.09.2016 um 04:36 schrieb Edward O'Callaghan: Save a few clocks by replacing calls to kmalloc followed by memcpy with a single call to kmemdup. Signed-off-by: Edward O'Callaghan <funfunc...@folklore1984.net> Reviewed-by: Christian König <christian.koe...@amd.com>. --- dri

Re: [PATCH 1/2] drm/amdgpu: add a new helper to free a BO in kernel allocations

ddress given. I think we should mirror that here or otherwise might run into problems calling kunmap() more often than kmap(). With that fixed both patches are Reviewed-by: Christian König <christian.koe...@amd.com>. Regards, Christian. + amdgpu_

[PATCH 2/2] drm/amdgpu: prevent command submission failures under memory pressure

From: Christian König <christian.koe...@amd.com> As last resort try to evict BOs from the current working set into other memory domains. This effectively prevents command submission failures when VM page tables have been swapped out. Signed-off-by: Christian König <christian.koe..

[PATCH 1/2] drm/amdgpu: only try again if we actually run into -ENOMEM

From: Christian König <christian.koe...@amd.com> All other errors can't be fixed by using a different memory domain. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 8 +++- 1 file changed, 3 insertions(+), 5 deletions(-)

Re: [PATCH] drm/amdgpu:implement CONTEXT_CONTROL (v3)

ursday, September 01, 2016 4:19 PM To: Liu, Monk <monk@amd.com> Cc: Christian König <deathsim...@vodafone.de>; amd-gfx@lists.freedesktop.org Subject: Re: [PATCH] drm/amdgpu:implement CONTEXT_CONTROL (v3) On Thu, Sep 1, 2016 at 9:37 AM, Liu, Monk <monk@amd.com> wrote: -Or

Re: [PATCH 2/2] drm/amdgpu: add gart recovery by gtt list

Am 30.08.2016 um 10:54 schrieb Chunming Zhou: Change-Id: I8daf687d43b51d69f9be36cae8b265deb88e5754 Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu.h| 15 +++ drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 +++

Re: [PATCH] drm/amdgpu: record error code when ring test failed

Am 30.08.2016 um 11:59 schrieb Chunming Zhou: Change-Id: I3a59f602a4d5ec42c8c184daa14eb8194b0dab9e Signed-off-by: Chunming Zhou <david1.z...@amd.com> Reviewed-by: Christian König <christian.koe...@amd.com>. --- drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | 5 +++-- 1 file changed,

Re: [PATCH 1/5] drm/amdgpu: add AMDGPU_GEM_CREATE_VRAM_LINEAR flag v2

Am 30.08.2016 um 03:49 schrieb Flora Cui: On Tue, Aug 30, 2016 at 09:57:56AM +0900, Michel Dänzer wrote: On 29/08/16 10:47 PM, Deucher, Alexander wrote: From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Christian König Sent: Monday, August 29, 2016 5:20 AM I still

Re: [PATCH 1/3] drm/amd/amdgpu: VCE ring test fail during S4 on CI

Am 30.08.2016 um 03:21 schrieb jimqu: Set up the VCE clock no matter dpm is enabled or not. NAK, if DPM or powerplay is enabled you are totally confusing it by messing with the clocks manually. Christian. Change-Id: I68e315f8b62c6e3a8636bc5e14036ecc11d980b4 Signed-off-by: JimQu

Re: [PATCH 1/2] drm/amdgpu: link all gtt when binding them

Am 30.08.2016 um 10:54 schrieb Chunming Zhou: Change-Id: I5dc3f9003e706d3a142aadb559bd85f6e0f2a03f Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu.h| 4 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 +++

Re: [PATCH 1/2] drm/ttm: remove unused placement flags

Am 09.09.2016 um 15:41 schrieb Deucher, Alexander: -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Christian König Sent: Friday, September 09, 2016 7:24 AM To: Zhang, Hawking; Koenig, Christian; Cui, Flora; Zhou, David(ChunMing); amd-gfx

[PATCH 3/3] drm/amdgpu: add a custom GTT memory manager

From: Christian König <christian.koe...@amd.com> Only allocate address space when we really need it. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/Makefile | 3 +- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 4 +- drivers

[PATCH 1/3] drm/amdgpu: add AMDGPU_BO_INVALID_OFFSET

From: Christian König <christian.koe...@amd.com> An invalid offset to be used by custom VRAM/GGT manager when we don't have an offset for an BO. Signed-off-by: Christian König <christian.koe...@amd.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 1 + drivers/gpu/dr

Re: [PATCH 1/2] drm/ttm: remove unused placement flags

Am 09.09.2016 um 15:54 schrieb Emil Velikov: On 9 September 2016 at 12:24, Christian König <deathsim...@vodafone.de> wrote: Hi Hawking, Removing the flag will make ttm_mem_type_from_place skip counting the corresponding placement and thus have impact on mem region create and bo mo

Re: [PATCH 0/3] rename file name.

Am 12.09.2016 um 10:36 schrieb Edward O'Callaghan: This series is, Reviewed-by: Edward O'Callaghan <funfunc...@folklore1984.net> Acked-by: Christian König <christian.koe...@amd.com> for the series. On 09/12/2016 06:22 PM, Rex Zhu wrote: As tonga/fiji/polaris10/11 use ppt tab

Re: [PATCH v2 00/16] powerplay code refactoring.

A few coding style issues, but nothing major. Taking patch #7 as an example, but most of this applies to the rest as well: +/*sviLoadLIneEn, SviLoadLineVddC, TDC_VDDC_ThrottleReleaseLimitPerc, TDC_MAWt, TdcWaterfallCtl, DTEAmbientTempBase, DisplayCac, BAPM_TEMP_GRADIENT */ That line looks

Re: [PATCH] drm/amdgpu: move some release handles into fail labels

Thanks for that info, I was busy most of the day bisecting this regression as well. Yeah, the bug is rather obvious, there is a "return 0;" in the good case missing here. So after loading the firmware we release it again. I will hack together a patch, just give me a minute. Christian. Am

Re: [PATCH 0/9] misc patches related to powerplay

} Best Regards Rex ---- *From:* Christian König <deathsim...@vodafone.de> *Sent:* Friday, September 9, 2016 10:23:26 PM *To:* Zhu, Rex; amd-gfx@lists.freedesktop.org *Subject:* Re: [PATCH 0/9] misc patches related to powerplay Patch #1-#7 and #9 are Acked-by: Christian Kö

<    1   2   3   4   5   6   7   8   9   10   >