Re: [PATCH] winsys/radeon: Allow visible VRAM size > 256MB with kernel driver >= 2.49

Thanks for fixing this Michel. Reviewed-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> On 01/31/2017 07:54 AM, Michel Dänzer wrote: From: Michel Dänzer <michel.daen...@amd.com> The kernel driver reports correct values now. Signed-off-by: Michel Dänzer <michel.daen...@am

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

recently. Fixes: "drm/amdgpu: improve AMDGPU_GEM_CREATE_VRAM_CLEARED handling (v2)" Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- Here's the report: https://hastebin.com/durodivoma.xml drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 4 +++- 1 file changed, 3 insertions(+

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

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. Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/am

[PATCH v2] drm/amdgpu: expose amdgpu_sensors on pre-powerplay chips

Totally untested but as long as read_sensor() has been recently implemented for dpm based boards, amdgpu_sensors can now be exposed. v2: - make sure read_sensor is not NULL on dpm chips - keep sanity check for powerplay chips Cc: Tom St Denis <tom.stde...@amd.com> Signed-off-by:

Re: [PATCH v2 2/2] drm/amdgpu: expose the current temperature and shader/memory clocks

On 02/14/2017 01:01 AM, Samuel Pitoiset wrote: The clocks are returned in Mhz and the temperature in millidegrees. v2: - add sub-queries for AMDPGU_INFO_GPU_SENSOR_* - do not break the ABI Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/

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

On 02/13/2017 07:58 PM, Nicolai Hähnle wrote: On 13.02.2017 19:38, Samuel Pitoiset wrote: On 02/13/2017 07:09 PM, Nicolai Hähnle wrote: On 13.02.2017 19:04, Nicolai Hähnle wrote: On 13.02.2017 18:49, Samuel Pitoiset wrote: On 02/13/2017 05:25 PM, Nicolai Hähnle wrote: On 09.02.2017 11

[PATCH 0/2] drm/amdgpu: expose temp and shader/memory clocks

the read_sensor() interface for pre-powerplay chips. The rest of the series is quite simple. This has only been tested on Polaris (RX480). Please review! Thanks. Samuel Pitoiset (2): drm/amdgpu: implement read_sensor() for pre-powerplay chips drm/amdgpu: expose the current temperature

[PATCH 1/2] drm/amdgpu: implement read_sensor() for pre-powerplay chips

Currently, only the GPU temperature, the shader clock and eventually the memory clock are implemented. The main goal is to expose this info to the userspace like Radeon. Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h | 4 +++- drive

[PATCH 2/2] drm/amdgpu: expose the current temperature and shader/memory clocks

The clocks are returned in Mhz and the temperature in millidegrees. Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 27 +++ include/uapi/drm/amdgpu

[PATCH v2 1/2] drm/amdgpu: implement read_sensor() for pre-powerplay chips

Add the GPU temperature, the shader clock and eventually the memory clock (as well as the GPU load on CI). The main goal is to expose this info to the userspace like Radeon. v2: - add AMDGPU_PP_SENSOR_GPU_LOAD on CI - update the commit description Signed-off-by: Samuel Pitoiset <samuel.pi

[PATCH v2 2/2] drm/amdgpu: expose the current temperature and shader/memory clocks

The clocks are returned in Mhz and the temperature in millidegrees. v2: - add sub-queries for AMDPGU_INFO_GPU_SENSOR_* - do not break the ABI Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++- drivers/gpu/drm/amd/

Re: [PATCH 1/2] drm/amdgpu: implement read_sensor() for pre-powerplay chips

patch yes. If you can swing that in here that would be helpful if not I can submit my own patch when this lands. Cheers, Tom On 02/13/2017 05:01 PM, Samuel Pitoiset wrote: Currently, only the GPU temperature, the shader clock and eventually the memory clock are implemented. The main goal

Re: [PATCH 1/2] drm/amdgpu: implement read_sensor() for pre-powerplay chips

On 02/14/2017 12:46 AM, Tom St Denis wrote: On 02/13/2017 06:40 PM, Samuel Pitoiset wrote: On 02/14/2017 12:17 AM, Tom St Denis wrote: Hi Samuel, Hi Tom, It would be helpful to modify amdgpu_debugfs_sensor_read() to support dpm based sensors as well. This will let me add it to umr

[PATCH] drm/amdgpu: expose amdgpu_sensors on pre-powerplay chips

Totally untested but as long as read_sensor() has been recently implemented for dpm based boards, amdgpu_sensors can now be exposed. Cc: Tom St Denis <tom.stde...@amd.com> Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 5 +

Re: [PATCH 0/2] drm/ttm, amdgpu: fix use-after-free in recent deadlock fix

On 02/14/2017 10:18 AM, Nicolai Hähnle wrote: Hi all, based on my current theory on how a deadlock could happen in the buffer allocation code, these two patches should fix the deadlock without having a use-after-free. I'm still working on a way to clean up the ttm_bo_init sequence overall,

Re: [PATCH 0/2] drm/ttm, amdgpu: fix use-after-free in recent deadlock fix

On 02/14/2017 10:18 AM, Nicolai Hähnle wrote: Hi all, based on my current theory on how a deadlock could happen in the buffer allocation code, these two patches should fix the deadlock without having a use-after-free. I'm still working on a way to clean up the ttm_bo_init sequence overall,

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

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. v2: fix computation Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drive

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

ty recently. v2: only release the mutex if resv is NULL Fixes: 12a852219583 ("drm/amdgpu: improve AMDGPU_GEM_CREATE_VRAM_CLEARED handling (v2)") Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 - 1 file changed, 4 insertio

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

On 02/09/2017 01:58 AM, Michel Dänzer wrote: On 09/02/17 08:35 AM, 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

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

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 <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 7 ++- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/d

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

This removes code duplication. Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 8 +--- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu

Re: [PATCH] drm/amdgpu: expose amdgpu_sensors on pre-powerplay chips

On 02/14/2017 06:01 PM, Deucher, Alexander wrote: -Original Message- From: StDenis, Tom Sent: Tuesday, February 14, 2017 11:58 AM To: Deucher, Alexander; 'Samuel Pitoiset'; amd-gfx@lists.freedesktop.org Subject: Re: [PATCH] drm/amdgpu: expose amdgpu_sensors on pre- powerplay chips

Re: [PATCH v2] drm/amdgpu: expose amdgpu_sensors on pre-powerplay chips

On 02/14/2017 05:41 PM, Tom St Denis wrote: On 14/02/17 10:08 AM, Samuel Pitoiset wrote: Totally untested but as long as read_sensor() has been recently implemented for dpm based boards, amdgpu_sensors can now be exposed. v2: - make sure read_sensor is not NULL on dpm chips - keep sanity

Re: [PATCH] drm/amdgpu: expose amdgpu_sensors on pre-powerplay chips

On 02/14/2017 03:04 PM, Tom St Denis wrote: On 13/02/17 07:23 PM, Samuel Pitoiset wrote: Totally untested but as long as read_sensor() has been recently implemented for dpm based boards, amdgpu_sensors can now be exposed. Cc: Tom St Denis <tom.stde...@amd.com> Signed-off-by: Samuel Pi

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

On 02/13/2017 05:28 PM, Nicolai Hähnle wrote: 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 v2 1/2] drm/amdgpu: fix a potential deadlock in amdgpu_bo_create_restricted()

On 02/13/2017 06:49 PM, Samuel Pitoiset wrote: On 02/13/2017 05:25 PM, Nicolai Hähnle wrote: On 09.02.2017 11:33, Samuel Pitoiset wrote: When ttm_bo_init() fails, the reservation mutex should be unlocked. In debug build, the kernel reported "possible recursive locking det

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

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 09.02.2017 11:33, Samuel Pitoiset wrote: When ttm_bo_init() fails, the reservation mutex should be unlocked. In debug build, the kernel reported

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

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 09.02.2017 11:33, Samuel Pitoiset wrote: When

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

On 02/13/2017 07:09 PM, Nicolai Hähnle wrote: On 13.02.2017 19:04, Nicolai Hähnle wrote: On 13.02.2017 18:49, Samuel Pitoiset wrote: On 02/13/2017 05:25 PM, Nicolai Hähnle wrote: On 09.02.2017 11:33, Samuel Pitoiset wrote: When ttm_bo_init() fails, the reservation mutex should

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

On 02/13/2017 07:41 PM, Christian König wrote: 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

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

take the BO as parameter, how can this happen? Regards, Christian. Am 09.02.2017 um 23:28 schrieb Samuel Pitoiset: 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 <samuel.pitoi...@gmail.

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

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 <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/

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

On 02/10/2017 11:19 AM, Christian König wrote: 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

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

On 02/10/2017 11:37 AM, Christian König wrote: Am 10.02.2017 um 11:22 schrieb Samuel Pitoiset: On 02/10/2017 11:19 AM, Christian König wrote: 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

Re: [PATCH] drm/amdgpu: expose amdgpu_sensors on pre-powerplay chips

On 02/15/2017 04:01 PM, Tom St Denis wrote: On 14/02/17 09:56 PM, Deucher, Alexander wrote: -Original Message- From: Michel Dänzer [mailto:mic...@daenzer.net] Sent: Tuesday, February 14, 2017 9:52 PM To: Deucher, Alexander; 'Samuel Pitoiset'; StDenis, Tom Cc: amd-gfx

Re: [PATCH v3 2/2] drm/amdgpu: expose GPU sensor related information

On 02/15/2017 08:15 PM, Samuel Pitoiset wrote: This includes shader/memory clocks, temperature, GPU load, etc. v2: - add sub-queries for AMDPGU_INFO_GPU_SENSOR_* - do not break the ABI v3: - return -ENOENT when amdgpu_dpm == 0 - expose more sensor queries Signed-off-by: Samuel

[PATCH v3] drm/amdgpu: expose amdgpu_sensors on pre-powerplay chips

ed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 6 ++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 6f021e70f15f..d63c44383660 100644 --- a/dri

[PATCH v4 2/2] drm/amdgpu: expose GPU sensor related information

- fix amdgpu_dpm check Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 +- drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 75 + include/uapi/drm/amdgpu_drm.h | 20 + 3 files chang

Re: Plan: BO move throttling for visible VRAM evictions

On 03/24/2017 05:33 PM, Marek Olšák wrote: Hi, I'm sharing this idea here, because it's something that has been decreasing our performance a lot recently, for example: http://openbenchmarking.org/prospect/1703011-RI-RADEONDIR06/7b7668cfc109d1c3dc27e871c8aea71ca13f23fa I think the problem

Re: [PATCH libdrm 1/3] headers: sync amdgpu_drm.h from airlied/drm-next

On 04/03/2017 10:27 AM, Nicolai Hähnle wrote: From: Nicolai Hähnle Changes include: PRT and preemption flags, sensor info, and some more changes for Vega10. Cool, thanks. I will be able to re-send the sensor info series for mesa. :) Generated using make

Re: [PATCH libdrm] amdgpu: allow to query GPU sensor related information

+02:00 Samuel Pitoiset <samuel.pitoi...@gmail.com>: This exposes amdgpu_query_sensor_info(). Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- amdgpu/amdgpu.h | 18 ++ amdgpu/amdgpu_gpu_info.c | 15 +++ 2 files changed, 33 insertions(+)

[PATCH libdrm v2] amdgpu: allow to query GPU sensor related information

This exposes amdgpu_query_sensor_info(). v2: - add amdgpu_query_sensor_info() to the symbols list Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- amdgpu/amdgpu-symbol-check | 1 + amdgpu/amdgpu.h| 18 ++ amdgpu/amdgpu_gpu_info.c

[PATCH libdrm] amdgpu: allow to query GPU sensor related information

This exposes amdgpu_query_sensor_info(). Signed-off-by: Samuel Pitoiset <samuel.pitoi...@gmail.com> --- amdgpu/amdgpu.h | 18 ++ amdgpu/amdgpu_gpu_info.c | 15 +++ 2 files changed, 33 insertions(+) diff --git a/amdgpu/amdgpu.h b/amdgpu/amdgpu.h

Re: [Mesa-dev] [RFC PATCH] radeonsi: set a per-buffer flag that disables inter-process sharing (v2)

On 07/20/2017 04:20 AM, zhoucm1 wrote: On 2017年07月19日 23:34, Marek Olšák wrote: On Jul 19, 2017 3:36 AM, "zhoucm1" > wrote: On 2017年07月19日 04:08, Marek Olšák wrote: From: Marek Olšák

Re: [PATCH] amdgpu: Add R600_DEBUG flag to reserve VMID per ctx.

On 10/31/2017 04:40 PM, Andrey Grodzovsky wrote: Signed-off-by: Andrey Grodzovsky --- configure.ac | 2 +- src/gallium/drivers/radeon/r600_pipe_common.c | 1 + src/gallium/drivers/radeon/r600_pipe_common.h | 1 +

[Vega10] [bisected] Performance regression with Dawn Of War III and RADV

Hi there, I found a performance regression that affects my Vega10 with Dawn Of War III and RADV (ie. 74FPS -> 71FPS). I bisected to: commit ea7f492b6ac6eced0d09ccb2b0a2909a0025e22e Author: Evan Quan Date: Mon Jul 16 17:23:19 2018 +0800 drm/amd/powerplay: correct the argument for

[PATCH] drm/amdgpu: print an error when the parser can't be initialized

Similar to other error messages, might help for tracking down issues. Signed-off-by: Samuel Pitoiset --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c

Re: [PATCH 6/8] drm/amdgpu: meld together VM fragment and huge page handling

, Christian. Am 08.11.18 um 17:17 schrieb Samuel Pitoiset: Hi, This change introduces GPU hangs with F1 2017 and Rise of Tomb Raider on RADV/GFX9. I bisected the issue. Can you have a look? Thanks! On 9/12/18 10:54 AM, Christian König wrote: This optimizes the generating of PTEs by walking

Re: [PATCH] drm/amdgpu: fix huge page handling on Vega10

On 11/13/18 9:14 AM, Christian König wrote: Am 12.11.18 um 22:45 schrieb Samuel Pitoiset: On 11/12/18 6:16 PM, Alex Deucher wrote: On Mon, Nov 12, 2018 at 12:09 PM Christian König wrote: We accidentially set the huge flag on the parent instead of the childs. This caused some VM faults

Re: [PATCH 6/8] drm/amdgpu: meld together VM fragment and huge page handling

Hi, This change introduces GPU hangs with F1 2017 and Rise of Tomb Raider on RADV/GFX9. I bisected the issue. Can you have a look? Thanks! On 9/12/18 10:54 AM, Christian König wrote: This optimizes the generating of PTEs by walking the hierarchy only once for a range and making changes as

Re: [PATCH] drm/amdgpu: fix huge page handling on Vega10

On 11/12/18 6:16 PM, Alex Deucher wrote: On Mon, Nov 12, 2018 at 12:09 PM Christian König wrote: We accidentially set the huge flag on the parent instead of the childs. This caused some VM faults under memory pressure. Signed-off-by: Christian König Yes, this fixes GPU hangs with F12017

Re: [RFC PATCH] drm/amdgpu: add support for user trap handlers

On 8/24/20 11:32 PM, Alex Deucher wrote: On Mon, Aug 24, 2020 at 2:33 PM Alex Deucher wrote: On Mon, Aug 24, 2020 at 7:57 AM Samuel Pitoiset wrote: A trap handler can be used by userspace to catch shader exceptions like divide by zero, memory violations etc. On GFX6-GFX8, the registers

Re: [RFC PATCH] drm/amdgpu: add support for user trap handlers

On 8/24/20 8:17 PM, Marek Olšák wrote: SET_SH_REG won't work with CP register shadowing. You need to use WRITE_DATA or WREG32. You are right, will fix. Marek On Mon, Aug 24, 2020 at 7:57 AM Samuel Pitoiset mailto:samuel.pitoi...@gmail.com>> wrote: A trap handler can b

[PATCH v2] drm/amdgpu: add support for user trap handlers

) Signed-off-by: Samuel Pitoiset --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 31 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h | 4 +++ drivers/gpu/drm/amd/amdgpu/amdgpu_job.h | 4 +++ drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h

Re: [PATCH v2] drm/amdgpu: add support for user trap handlers

On 8/28/20 9:57 AM, Christian König wrote: Am 25.08.20 um 16:07 schrieb Samuel Pitoiset: A trap handler can be used by userspace to catch shader exceptions like divide by zero, memory violations etc. On GFX6-GFX8, the registers used to configure TBA/TMA aren't privileged and can be configured

Re: [PATCH v2] drm/amdgpu: add support for user trap handlers

On 8/28/20 10:23 AM, Christian König wrote: Am 28.08.20 um 10:14 schrieb Samuel Pitoiset: On 8/28/20 9:57 AM, Christian König wrote: Am 25.08.20 um 16:07 schrieb Samuel Pitoiset: A trap handler can be used by userspace to catch shader exceptions like divide by zero, memory violations etc

[RFC PATCH] drm/amdgpu: add support for user trap handlers

CS chunk that can be used to set the TBA/TMA virtual address at submit time. TODO: - add GFX 6,7 and 10 support - rebase on top of amd-staging-drm-next (this branch currently hangs my GPU at boot) Signed-off-by: Samuel Pitoiset --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 31

Re: [PATCH v2] drm/amdgpu: add support for user trap handlers

On 8/28/20 10:23 AM, Christian König wrote: Am 28.08.20 um 10:14 schrieb Samuel Pitoiset: On 8/28/20 9:57 AM, Christian König wrote: Am 25.08.20 um 16:07 schrieb Samuel Pitoiset: A trap handler can be used by userspace to catch shader exceptions like divide by zero, memory violations etc

[PATCH v3] drm/amdgpu: add support for user trap handlers

. At the moment, we don't know how to set them via the CP, so they are only emitted if a VMID is reserved. This introduces a new CS chunk that can be used to set the TBA/TMA virtual address at submit time. Signed-off-by: Samuel Pitoiset --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 31

Re: [PATCH v3] drm/amdgpu: add support for user trap handlers

Added GF10.3 support. I re-tested this on GFX9 and it still works, though on GFX10+ the trap handler is never reached, is there something obviously wrong in this patch? Thanks! On 5/6/21 8:54 AM, Samuel Pitoiset wrote: A trap handler can be used by userspace to catch shader exceptions like

Re: [PATCH] drm/amdgpu: add missing NULL check

I can confirm this patch fixes the kernel crash I reported. But as discussed with Christian, we should find the root cause. On 10/6/23 14:11, Christian König wrote: bo->tbo.resource can easily be NULL here. Signed-off-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_object.h | 2

[PATCH] drm/amdgpu: fix clearing mappings for BOs that are always valid in VM

-off-by: Samuel Pitoiset --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 1 file changed, 12 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 143d11afe0e5..eff73c428b12 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c

[PATCH v2] drm/amdgpu: fix clearing mappings for BOs that are always valid in VM

are correctly handled there. Cc: sta...@vger.kernel.org Signed-off-by: Samuel Pitoiset Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 1 file changed, 12 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu