From: Sasha Levin <sashal@kernel.org>
To: stable@vger.kernel.org, linux-kernel@vger.kernel.org
Cc: Yong Zhao <yong.zhao@amd.com>,
Felix Kuehling <Felix.Kuehling@amd.com>,
Alex Deucher <alexander.deucher@amd.com>,
Sasha Levin <alexander.levin@microsoft.com>
Subject: [PATCH AUTOSEL 4.18 45/48] drm/amdkfd: Change the control stack MTYPE from UC to NC on GFX9
Date: Fri, 5 Oct 2018 12:14:21 -0400 [thread overview]
Message-ID: <20181005161424.20521-45-sashal@kernel.org> (raw)
In-Reply-To: <20181005161424.20521-1-sashal@kernel.org>
From: Yong Zhao <yong.zhao@amd.com>
[ Upstream commit 15426dbb65c5b37680d27e84d58a1ed3b8532518 ]
CWSR fails on Raven if the control stack is MTYPE_UC, which is used
for regular GART mappings. As a workaround we map it using MTYPE_NC.
The MEC firmware expects the control stack at one page offset from the
start of the MQD so it is part of the MQD allocation on GFXv9. AMDGPU
added a memory allocation flag just for this purpose.
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <alexander.levin@microsoft.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 6 +++++-
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 2 +-
drivers/gpu/drm/amd/amdkfd/kfd_device.c | 3 ++-
drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c | 2 +-
drivers/gpu/drm/amd/include/kgd_kfd_interface.h | 2 +-
5 files changed, 10 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index 305143fcc1ce..1ac7933cccc5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -245,7 +245,7 @@ int amdgpu_amdkfd_resume(struct amdgpu_device *adev)
int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
void **mem_obj, uint64_t *gpu_addr,
- void **cpu_ptr)
+ void **cpu_ptr, bool mqd_gfx9)
{
struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
struct amdgpu_bo *bo = NULL;
@@ -261,6 +261,10 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
bp.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC;
bp.type = ttm_bo_type_kernel;
bp.resv = NULL;
+
+ if (mqd_gfx9)
+ bp.flags |= AMDGPU_GEM_CREATE_MQD_GFX9;
+
r = amdgpu_bo_create(adev, &bp, &bo);
if (r) {
dev_err(adev->dev,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index a8418a3f4e9d..e3cf1c9fb3db 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -129,7 +129,7 @@ bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid);
/* Shared API */
int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
void **mem_obj, uint64_t *gpu_addr,
- void **cpu_ptr);
+ void **cpu_ptr, bool mqd_gfx9);
void free_gtt_mem(struct kgd_dev *kgd, void *mem_obj);
void get_local_mem_info(struct kgd_dev *kgd,
struct kfd_local_mem_info *mem_info);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 7ee6cec2c060..6881b5a9275f 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -423,7 +423,8 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
if (kfd->kfd2kgd->init_gtt_mem_allocation(
kfd->kgd, size, &kfd->gtt_mem,
- &kfd->gtt_start_gpu_addr, &kfd->gtt_start_cpu_ptr)){
+ &kfd->gtt_start_gpu_addr, &kfd->gtt_start_cpu_ptr,
+ false)) {
dev_err(kfd_device, "Could not allocate %d bytes\n", size);
goto out;
}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
index 684054ff02cd..8da079cc6fb9 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
@@ -63,7 +63,7 @@ static int init_mqd(struct mqd_manager *mm, void **mqd,
ALIGN(sizeof(struct v9_mqd), PAGE_SIZE),
&((*mqd_mem_obj)->gtt_mem),
&((*mqd_mem_obj)->gpu_addr),
- (void *)&((*mqd_mem_obj)->cpu_ptr));
+ (void *)&((*mqd_mem_obj)->cpu_ptr), true);
} else
retval = kfd_gtt_sa_allocate(mm->dev, sizeof(struct v9_mqd),
mqd_mem_obj);
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
index 5733fbee07f7..f56b7553e5ed 100644
--- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -266,7 +266,7 @@ struct tile_config {
struct kfd2kgd_calls {
int (*init_gtt_mem_allocation)(struct kgd_dev *kgd, size_t size,
void **mem_obj, uint64_t *gpu_addr,
- void **cpu_ptr);
+ void **cpu_ptr, bool mqd_gfx9);
void (*free_gtt_mem)(struct kgd_dev *kgd, void *mem_obj);
--
2.17.1
next prev parent reply other threads:[~2018-10-05 16:15 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-05 16:13 [PATCH AUTOSEL 4.18 01/48] ASoC: dapm: Fix NULL pointer deference on CODEC to CODEC DAIs Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 02/48] ASoC: max98373: Added speaker FS gain cotnrol register to volatile Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 03/48] ASoC: rt5514: Fix the issue of the delay volume applied again Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 04/48] selftests: android: move config up a level Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 05/48] selftests: kselftest: Remove outdated comment Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 06/48] ASoC: max98373: Added 10ms sleep after amp software reset Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 07/48] ASoC: wm8804: Add ACPI support Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 08/48] ASoC: sigmadsp: safeload should not have lower byte limit Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 09/48] ASoC: q6routing: initialize data correctly Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 10/48] selftests: add headers_install to lib.mk Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 11/48] selftests/efivarfs: add required kernel configs Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 12/48] selftests: memory-hotplug: add required configs Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 13/48] ASoC: rsnd: adg: care clock-frequency size Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 14/48] ASoC: rsnd: don't fallback to PIO mode when -EPROBE_DEFER Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 15/48] hwmon: (nct6775) Fix access to fan pulse registers Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 16/48] Fix cg_read_strcmp() Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 17/48] Add tests for memory.oom.group Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 18/48] ASoC: AMD: Ensure reset bit is cleared before configuring Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 19/48] drm/pl111: Make sure of_device_id tables are NULL terminated Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 20/48] Bluetooth: SMP: Fix trying to use non-existent local OOB data Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 21/48] Bluetooth: Use correct tfm to generate " Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 22/48] Bluetooth: hci_ldisc: Free rw_semaphore on close Sasha Levin
2018-10-05 16:13 ` [PATCH AUTOSEL 4.18 23/48] mfd: omap-usb-host: Fix dts probe of children Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 24/48] KVM: PPC: Book3S HV: Don't use compound_order to determine host mapping size Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 25/48] scsi: iscsi: target: Don't use stack buffer for scatterlist Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 26/48] scsi: qla2xxx: Fix an endian bug in fcpcmd_is_corrupted() Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 27/48] sound: enable interrupt after dma buffer initialization Sasha Levin
2018-10-08 9:34 ` Mark Brown
2018-10-08 9:36 ` Takashi Iwai
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 28/48] sound: don't call skl_init_chip() to reset intel skl soc Sasha Levin
2018-10-08 9:34 ` Mark Brown
2018-10-08 9:37 ` Takashi Iwai
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 29/48] bpf: btf: Fix end boundary calculation for type section Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 30/48] bpf: use __GFP_COMP while allocating page Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 31/48] hwmon: (nct6775) Fix virtual temperature sources for NCT6796D Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 32/48] hwmon: (nct6775) Fix RPM output for fan7 on NCT6796D Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 33/48] stmmac: fix valid numbers of unicast filter entries Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 34/48] hwmon: (nct6775) Use different register to get fan RPM for fan7 Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 35/48] net: ethernet: ti: add missing GENERIC_ALLOCATOR dependency Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 36/48] net: macb: disable scatter-gather for macb on sama5d3 Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 37/48] ARM: dts: at91: add new compatibility string " Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 38/48] PCI: hv: support reporting serial number as slot information Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 39/48] hv_netvsc: pair VF based on serial number Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 40/48] clk: x86: add "ether_clk" alias for Bay Trail / Cherry Trail Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 41/48] clk: x86: Stop marking clocks as CLK_IS_CRITICAL Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 42/48] pinctrl: cannonlake: Fix gpio base for GPP-E Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 43/48] x86/kvm/lapic: always disable MMIO interface in x2APIC mode Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 44/48] drm/amdgpu: Fix SDMA HQD destroy error on gfx_v7 Sasha Levin
2018-10-05 16:14 ` Sasha Levin [this message]
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 46/48] drm/amdkfd: Fix ATS capablity was not reported correctly on some APUs Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 47/48] ubifs: Check for name being NULL while mounting Sasha Levin
2018-10-05 16:14 ` [PATCH AUTOSEL 4.18 48/48] mm: slowly shrink slabs with a relatively small number of objects Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20181005161424.20521-45-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=Felix.Kuehling@amd.com \
--cc=alexander.deucher@amd.com \
--cc=alexander.levin@microsoft.com \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
--cc=yong.zhao@amd.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox