Re: [PATCH 1/9] drm/amdkfd: Avoid name confusion involved in queue unmapping

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > From: Yong Zhao > > When unmapping the queues from HW scheduler, there are two actions: > reset and preempt. So naming the variables with only preempt is > inapproriate. > > For functions such as destroy_queues_cpsch, what they do actually

Re: [PATCH 0/9] KFD queue management improvements and fixes

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > These patches cherry-pick most fixes, improvements and refactoring in > the device queue manager from our internal branch. I squashed some > patches that applied incremental fixes to the same code. Patch 5 is > the extreme case that comes wi

Re: [PATCH 2/9] drm/amdkfd: Simplify execute and unmap queue APIs

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > From: Yong Zhao > > Pass filter parameters directly to unmap_queues_cpsch. > > Also remove lock parameters and do locking explicitly in the caller. > > Signed-off-by: Yong Zhao > Signed-off-by: Felix Kuehling > --- > .../gpu/drm/amd/amdk

Re: [PATCH 4/9] drm/amdkfd: Avoid submitting an unnecessary packet to HWS

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > From: Yong Zhao > > Signed-off-by: Yong Zhao > Signed-off-by: Felix Kuehling > --- > .../gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 28 > -- > 1 file changed, 16 insertions(+), 12 deletions(-) > > diff --git a/

Re: [PATCH 3/9] drm/amdkfd: Fix MQD updates

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > When a queue is mapped, the MQD is owned by the FW. The FW overwrites > the MQD on the next unmap operation. Therefore the queue must be > unmapped before updating the MQD. > > For the non-HWS case, also fix disabling of queues and creation

Re: [PATCH 5/9] drm/amdkfd: Improve process termination handling in the queue manager

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > Separate device queue termination from process queue manager > termination. Unmap all queues at once instead of one at a time. > Unmap device queues before the PASID is unbound, in the > kfd_process_iommu_unbind_callback. > > When resetting

Re: [PATCH 6/9] drm/amdkfd: Compress unnecessary function parameters

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > From: Yong Zhao > > Signed-off-by: Yong Zhao > --- > drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 3 +-- > drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c| 5 +++-- > drivers/gpu/drm/amd/amdkfd/kfd_priv.h

Re: [PATCH 7/9] drm/amdkfd: Clean up process queue management

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > Removed unused num_concurrent_processes. > > Implemented counting of queues in QPD. This makes counting the queue > list repeatedly in several places unnecessary. > > Signed-off-by: Felix Kuehling > --- > drivers/gpu/drm/amd/amdkfd/kfd_dev

Re: [PATCH 8/9] drm/amdkfd: Limit queue number per process and device to 127

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > HWS uses bit 7 in the queue number of the map process packet for an > undocumented feature. Therefore the queue number per process and > device must be 127 or less. > > Signed-off-by: Felix Kuehling > --- > drivers/gpu/drm/amd/amdkfd/kfd_p

Re: [PATCH 9/9] drm/amdkfd: Improve multiple SDMA queues support per process

2017-10-08 Thread Oded Gabbay
On Wed, Sep 27, 2017 at 7:09 AM, Felix Kuehling wrote: > From: shaoyunl > > HWS does not support over-subscription and the scheduler can not internally > modify the engine. Driver needs to program the correct engine ID. > > Fix the queue and engine selection to create queues on alternating SDMA >

RE: [PATCH] drm/amd/powerplay: get the temperature on CZ

2017-10-08 Thread Deucher, Alexander
> -Original Message- > From: Sahu, Satyajit > Sent: Friday, October 06, 2017 2:28 AM > To: amd-gfx@lists.freedesktop.org > Cc: Deucher, Alexander; brahma_sw_dev > Subject: [PATCH] drm/amd/powerplay: get the temperature on CZ > > From: Satyajit Sahu > > Setting the function pointer to the

Re: [amd-staging-drm-next] regression - no fan info (sensors) on RX580

2017-10-08 Thread Dieter Nützel
Sorry Rex, after return from our vacation, I've tested latest amd-staging-drm-next (e5f6a57e350a) but it is NOT solved on my RX580. I'll try bisecting if I find some more time in the coming days. amdgpu-pci-0100 Adapter: PCI adapter temp1:+27.0°C (crit = +0.0°C, hyst = +0.0°C) 01:00.

[pull] radeon, amdgpu, and ttm drm-next-4.15

2017-10-08 Thread Alex Deucher
Hi Dave, More new stuff for 4.15. Highlights: - Add clock query interface for raven - Add new FENCE_TO_HANDLE ioctl - UVD video encode ring support on polaris - transparent huge page DMA support - deadlock fixes - compute pipe lru tweaks - powerplay cleanups and regression fixes - fix duplicate sy

Re: [amd-staging-drm-next] regression - no fan info (sensors) on RX580

2017-10-08 Thread Dieter Nützel
OK, got it but can't revert the commit clean. amdgpu-pci-0100 Adapter: PCI adapter fan1: 873 RPM temp1:+26.0°C (crit = +0.0°C, hyst = +0.0°C) SOURCE/amd-staging-drm-next> git bisect good 0944c350c8eddf4064e7abb881dd245032fdfa23 is the first bad commit commit 0944c350c8eddf4064

Re: [PATCH 1/2] drm/amdgpu:resolv deadlock between reset and cs_ioctl v4.

2017-10-08 Thread Chunming Zhou
you added the ctx mutex to keep the job pushing in order, that's good, Acked-by: Chunming Zhou BTW: after you added ctx mutex for cs, I think the thread lock in libdrm isn't need any more, we can remove it now. Regards, David Zhou On 2017年10月07日 02:20, Andrey Grodzovsky wrote: From: Mo

RE: [PATCH 12/12] drm/amdgpu/sriov:no shadow buffer recovery

2017-10-08 Thread Liu, Monk
Any updates for the rest patches ? From: Koenig, Christian Sent: 2017年10月4日 18:56 To: Liu, Monk ; amd-gfx@lists.freedesktop.org Subject: Re: [PATCH 12/12] drm/amdgpu/sriov:no shadow buffer recovery Ah! Sorry, my fault. I've missed the "no" and thought you wanted to abandon all processing because

[PATCH 5/9] drm/amdgpu: add new pp function point notify_smu_memory_info

2017-10-08 Thread Rex Zhu
Change-Id: Ib915cd5ac32a6b75667ff6e8ddafcea61e06cd67 Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h| 6 ++ drivers/gpu/drm/amd/include/kgd_pp_interface.h | 5 + 2 files changed, 11 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/g

[PATCH 1/9] drm/amd/powerplay: export new smu messages for vega10

2017-10-08 Thread Rex Zhu
Change-Id: I5afa09980174ef191c7aa4e4f9dadaaa189783ca Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h index cb070eb

[PATCH 6/9] drm/amd/powerplay: implement notify_smu_memory_info on Powerplay

2017-10-08 Thread Rex Zhu
Change-Id: I69a8a7637be35b9b03d7f7390cd3571cd24f9781 Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 36 +++ 1 file changed, 36 insertions(+) diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerpl

[PATCH 3/9] drm/amd/powrplay: implement function notify_cac_buffer_info on Vega

2017-10-08 Thread Rex Zhu
Change-Id: I4d50bf04ba6f5caf6919b6177517c7b38b9a606a Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 28 ++ 1 file changed, 28 insertions(+) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/

[PATCH 4/9] drm/amd/powerplay: implement function notify_cac_buffer_info on VI

2017-10-08 Thread Rex Zhu
Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c | 28 +++ drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 42 ++ drivers/gpu/drm/amd/powerplay/inc/smumgr.h | 5 +++ drivers/gpu/drm/amd/powerplay/smumgr/ci_smc.c | 10

[PATCH 7/9] drm/amdgpu: add smu_memory_pool_size module parameter

2017-10-08 Thread Rex Zhu
this allow allocate system memmoy for smu debug usage. Change-Id: Iac0489e528395448abeaf23a22cd6a1031a5b55b Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 5 + 2 files changed, 6 insertions(+) diff --git a/drivers/gpu/drm/

[PATCH 8/9] drm/amdgpu: allocate requested gtt buffer for smu

2017-10-08 Thread Rex Zhu
v2: simplify check smu_memory_size code. simplify allocate smu memroy code. Change-Id: I8eb4f542dc2351c6393e4723f4985df92ff527cd Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu.h| 2 + drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 79 ++ drive

[PATCH 9/9] drm/amd/powerplay: delete an outdated comment in amd_powerplay.c

2017-10-08 Thread Rex Zhu
Change-Id: I96abfea4eb7d257d2b6461392683c280b268ef76 Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index c7ab8b3..6b0cf8e

[PATCH 2/9] drm/amd/powerplay: add new function point in hwmgr.

2017-10-08 Thread Rex Zhu
used for notify SMU the allocated buffer address. Change-Id: I91badca7729b8d9c35faf7fc09dbdee70c26099a Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/powerplay/inc/hwmgr.h | 6 ++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/p

[PATCH] drm/amdgpu: refine code delete duplicated error handling

2017-10-08 Thread Rex Zhu
in function amdgpu_ucode_init_bo, when failed, it will set load_type to AMDGPU_FW_LOAD_DIRECT. Change-Id: Id671b168425dc46331aaf4bd2b734e88b9bfaf3c Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/driver

RE: [PATCH 2/2] dmr/amdgpu: Move old fence waiting before reservation lock is aquired.

2017-10-08 Thread Liu, Monk
Hi Andrey Where is your first patch (1/2)? I only saw the 2/2 BR Monk -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Andrey Grodzovsky Sent: 2017年10月7日 2:20 To: amd-gfx@lists.freedesktop.org Cc: Grodzovsky, Andrey ; Koenig, Christian Subj

RE: [PATCH 5/5] drm/amd/sched: signal and free remaining fences in amd_sched_entity_fini

2017-10-08 Thread Liu, Monk
Christian > It would be really nice to have an error code set on s_fence->finished before > it is signaled, use dma_fence_set_error() for this. For gpu reset patches (already submitted to pub) I would make kernel return -ENODEV if the waiting fence (in cs_wait or wait_fences IOCTL) founded as

Re: [PATCH] drm/amdgpu: Set the correct value for PDEs/PTEs of ATC memory on Raven

2017-10-08 Thread Zhang, Jerry (Junwei)
On 10/06/2017 11:35 PM, Yong Zhao wrote: From: Yong Zhao Without the additional bits set in PDEs/PTEs, the ATC memory access would have failed on Raven. Change-Id: I28429ef6d39cdb01dc6f17fea4264ee22d7121d4 Signed-off-by: Yong Zhao Acked-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgp