* [PATCH v3 01/10] dt-bindings: display/msm/gmu: Add Adreno 7[34]0 GMU
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
@ 2023-08-23 12:55 ` Konrad Dybcio
2023-08-23 12:55 ` [PATCH v3 02/10] dt-bindings: display/msm/gmu: Allow passing QMP handle Konrad Dybcio
` (8 subsequent siblings)
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:55 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Krzysztof Kozlowski, Neil Armstrong
The GMU on the A7xx series is pretty much the same as on the A6xx parts.
It's now "smarter", needs a bit less register writes and controls more
things (like inter-frame power collapse) mostly internally (instead of
us having to write to G[PM]U_[CG]X registers from APPS)
The only difference worth mentioning is the now-required DEMET clock,
which is strictly required for things like asserting reset lines, not
turning it on results in GMU not being fully functional (all OOB requests
would fail and HFI would hang after the first submitted OOB).
Describe the A730 and A740 GMU.
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
.../devicetree/bindings/display/msm/gmu.yaml | 40 +++++++++++++++++++++-
1 file changed, 39 insertions(+), 1 deletion(-)
diff --git a/Documentation/devicetree/bindings/display/msm/gmu.yaml b/Documentation/devicetree/bindings/display/msm/gmu.yaml
index 5fc4106110ad..20ddb89a4500 100644
--- a/Documentation/devicetree/bindings/display/msm/gmu.yaml
+++ b/Documentation/devicetree/bindings/display/msm/gmu.yaml
@@ -21,7 +21,7 @@ properties:
compatible:
oneOf:
- items:
- - pattern: '^qcom,adreno-gmu-6[0-9][0-9]\.[0-9]$'
+ - pattern: '^qcom,adreno-gmu-[67][0-9][0-9]\.[0-9]$'
- const: qcom,adreno-gmu
- const: qcom,adreno-gmu-wrapper
@@ -213,6 +213,44 @@ allOf:
- const: axi
- const: memnoc
+ - if:
+ properties:
+ compatible:
+ contains:
+ enum:
+ - qcom,adreno-gmu-730.1
+ - qcom,adreno-gmu-740.1
+ then:
+ properties:
+ reg:
+ items:
+ - description: Core GMU registers
+ - description: Resource controller registers
+ - description: GMU PDC registers
+ reg-names:
+ items:
+ - const: gmu
+ - const: rscc
+ - const: gmu_pdc
+ clocks:
+ items:
+ - description: GPU AHB clock
+ - description: GMU clock
+ - description: GPU CX clock
+ - description: GPU AXI clock
+ - description: GPU MEMNOC clock
+ - description: GMU HUB clock
+ - description: GPUSS DEMET clock
+ clock-names:
+ items:
+ - const: ahb
+ - const: gmu
+ - const: cxo
+ - const: axi
+ - const: memnoc
+ - const: hub
+ - const: demet
+
- if:
properties:
compatible:
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH v3 02/10] dt-bindings: display/msm/gmu: Allow passing QMP handle
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
2023-08-23 12:55 ` [PATCH v3 01/10] dt-bindings: display/msm/gmu: Add Adreno 7[34]0 GMU Konrad Dybcio
@ 2023-08-23 12:55 ` Konrad Dybcio
2023-08-23 12:55 ` [PATCH v3 03/10] dt-bindings: display/msm/gpu: Allow A7xx SKUs Konrad Dybcio
` (7 subsequent siblings)
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:55 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong, Krzysztof Kozlowski
When booting the GMU, the QMP mailbox should be pinged about some tunables
(e.g. adaptive clock distribution state). To achieve that, a reference to
it is necessary. Allow it and require it with A730.
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
Documentation/devicetree/bindings/display/msm/gmu.yaml | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/Documentation/devicetree/bindings/display/msm/gmu.yaml b/Documentation/devicetree/bindings/display/msm/gmu.yaml
index 20ddb89a4500..e132dbff3c4a 100644
--- a/Documentation/devicetree/bindings/display/msm/gmu.yaml
+++ b/Documentation/devicetree/bindings/display/msm/gmu.yaml
@@ -64,6 +64,10 @@ properties:
iommus:
maxItems: 1
+ qcom,qmp:
+ $ref: /schemas/types.yaml#/definitions/phandle
+ description: Reference to the AOSS side-channel message RAM
+
operating-points-v2: true
opp-table:
@@ -251,6 +255,9 @@ allOf:
- const: hub
- const: demet
+ required:
+ - qcom,qmp
+
- if:
properties:
compatible:
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH v3 03/10] dt-bindings: display/msm/gpu: Allow A7xx SKUs
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
2023-08-23 12:55 ` [PATCH v3 01/10] dt-bindings: display/msm/gmu: Add Adreno 7[34]0 GMU Konrad Dybcio
2023-08-23 12:55 ` [PATCH v3 02/10] dt-bindings: display/msm/gmu: Allow passing QMP handle Konrad Dybcio
@ 2023-08-23 12:55 ` Konrad Dybcio
2023-08-23 12:55 ` [PATCH v3 04/10] drm/msm/a6xx: Add missing regs for A7XX Konrad Dybcio
` (6 subsequent siblings)
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:55 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Krzysztof Kozlowski, Neil Armstrong
Allow A7xx SKUs, such as the A730 GPU found on SM8450 and friends.
They use GMU for all things DVFS, just like most A6xx GPUs.
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
Documentation/devicetree/bindings/display/msm/gpu.yaml | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/Documentation/devicetree/bindings/display/msm/gpu.yaml b/Documentation/devicetree/bindings/display/msm/gpu.yaml
index 56b9b247e8c2..b019db954793 100644
--- a/Documentation/devicetree/bindings/display/msm/gpu.yaml
+++ b/Documentation/devicetree/bindings/display/msm/gpu.yaml
@@ -23,7 +23,7 @@ properties:
The driver is parsing the compat string for Adreno to
figure out the gpu-id and patch level.
items:
- - pattern: '^qcom,adreno-[3-6][0-9][0-9]\.[0-9]$'
+ - pattern: '^qcom,adreno-[3-7][0-9][0-9]\.[0-9]$'
- const: qcom,adreno
- description: |
The driver is parsing the compat string for Imageon to
@@ -203,7 +203,7 @@ allOf:
properties:
compatible:
contains:
- pattern: '^qcom,adreno-6[0-9][0-9]\.[0-9]$'
+ pattern: '^qcom,adreno-[67][0-9][0-9]\.[0-9]$'
then: # Starting with A6xx, the clocks are usually defined in the GMU node
properties:
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH v3 04/10] drm/msm/a6xx: Add missing regs for A7XX
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
` (2 preceding siblings ...)
2023-08-23 12:55 ` [PATCH v3 03/10] dt-bindings: display/msm/gpu: Allow A7xx SKUs Konrad Dybcio
@ 2023-08-23 12:55 ` Konrad Dybcio
2023-08-23 12:55 ` [PATCH v3 05/10] drm/msm/a6xx: Add skeleton A7xx support Konrad Dybcio
` (5 subsequent siblings)
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:55 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong
Add some missing definitions required for A7 support.
This may be substituted with a mesa header sync.
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
drivers/gpu/drm/msm/adreno/a6xx.xml.h | 9 +++++++++
drivers/gpu/drm/msm/adreno/a6xx_gmu.xml.h | 8 ++++++++
2 files changed, 17 insertions(+)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx.xml.h b/drivers/gpu/drm/msm/adreno/a6xx.xml.h
index 1c051535fd4a..863b5e3b0e67 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a6xx.xml.h
@@ -1114,6 +1114,12 @@ enum a6xx_tex_type {
#define REG_A6XX_CP_MISC_CNTL 0x00000840
#define REG_A6XX_CP_APRIV_CNTL 0x00000844
+#define A6XX_CP_APRIV_CNTL_CDWRITE 0x00000040
+#define A6XX_CP_APRIV_CNTL_CDREAD 0x00000020
+#define A6XX_CP_APRIV_CNTL_RBRPWB 0x00000008
+#define A6XX_CP_APRIV_CNTL_RBPRIVLEVEL 0x00000004
+#define A6XX_CP_APRIV_CNTL_RBFETCH 0x00000002
+#define A6XX_CP_APRIV_CNTL_ICACHE 0x00000001
#define REG_A6XX_CP_PREEMPT_THRESHOLD 0x000008c0
@@ -1939,6 +1945,8 @@ static inline uint32_t REG_A6XX_RBBM_PERFCTR_RBBM_SEL(uint32_t i0) { return 0x00
#define REG_A6XX_RBBM_CLOCK_HYST_TEX_FCHE 0x00000122
+#define REG_A7XX_RBBM_CLOCK_HYST2_VFD 0x0000012f
+
#define REG_A6XX_RBBM_LPAC_GBIF_CLIENT_QOS_CNTL 0x000005ff
#define REG_A6XX_DBGC_CFG_DBGBUS_SEL_A 0x00000600
@@ -8252,5 +8260,6 @@ static inline uint32_t A6XX_CX_DBGC_CFG_DBGBUS_BYTEL_1_BYTEL15(uint32_t val)
#define REG_A6XX_CX_MISC_SYSTEM_CACHE_CNTL_1 0x00000002
+#define REG_A7XX_CX_MISC_TCM_RET_CNTL 0x00000039
#endif /* A6XX_XML */
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.xml.h b/drivers/gpu/drm/msm/adreno/a6xx_gmu.xml.h
index fcd9eb53baf8..5b66efafc901 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.xml.h
@@ -360,6 +360,12 @@ static inline uint32_t A6XX_GMU_GPU_NAP_CTRL_SID(uint32_t val)
#define REG_A6XX_GMU_GENERAL_7 0x000051cc
+#define REG_A6XX_GMU_GENERAL_8 0x000051cd
+
+#define REG_A6XX_GMU_GENERAL_9 0x000051ce
+
+#define REG_A6XX_GMU_GENERAL_10 0x000051cf
+
#define REG_A6XX_GMU_ISENSE_CTRL 0x0000515d
#define REG_A6XX_GPU_CS_ENABLE_REG 0x00008920
@@ -471,6 +477,8 @@ static inline uint32_t A6XX_GMU_GPU_NAP_CTRL_SID(uint32_t val)
#define REG_A6XX_RSCC_SEQ_BUSY_DRV0 0x00000101
+#define REG_A7XX_RSCC_SEQ_MEM_0_DRV0_A740 0x00000154
+
#define REG_A6XX_RSCC_SEQ_MEM_0_DRV0 0x00000180
#define REG_A6XX_RSCC_TCS0_DRV0_STATUS 0x00000346
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH v3 05/10] drm/msm/a6xx: Add skeleton A7xx support
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
` (3 preceding siblings ...)
2023-08-23 12:55 ` [PATCH v3 04/10] drm/msm/a6xx: Add missing regs for A7XX Konrad Dybcio
@ 2023-08-23 12:55 ` Konrad Dybcio
2023-09-11 15:30 ` Dmitry Baryshkov
2023-08-23 12:55 ` [PATCH v3 06/10] drm/msm/a6xx: Send ACD state to QMP at GMU resume Konrad Dybcio
` (4 subsequent siblings)
9 siblings, 1 reply; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:55 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong
A7xx GPUs are - from kernel's POV anyway - basically another generation
of A6xx. They build upon the A650/A660_family advancements, skipping some
writes (presumably more values are preset correctly on reset), adding
some new ones and changing others.
One notable difference is the introduction of a second shadow, called BV.
To handle this with the current code, allocate it right after the current
RPTR shadow.
BV handling and .submit are mostly based on Jonathan Marek's work.
All A7xx GPUs are assumed to have a GMU.
A702 is not an A7xx-class GPU, it's a weird forked A610.
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 95 +++++--
drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 451 ++++++++++++++++++++++++++++----
drivers/gpu/drm/msm/adreno/adreno_gpu.c | 1 +
drivers/gpu/drm/msm/adreno/adreno_gpu.h | 10 +-
drivers/gpu/drm/msm/msm_ringbuffer.h | 2 +
5 files changed, 478 insertions(+), 81 deletions(-)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
index 03fa89bf3e4b..75984260898e 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
@@ -200,9 +200,10 @@ int a6xx_gmu_wait_for_idle(struct a6xx_gmu *gmu)
static int a6xx_gmu_start(struct a6xx_gmu *gmu)
{
+ struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
+ struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
+ u32 mask, reset_val, val;
int ret;
- u32 val;
- u32 mask, reset_val;
val = gmu_read(gmu, REG_A6XX_GMU_CM3_DTCM_START + 0xff8);
if (val <= 0x20010004) {
@@ -218,7 +219,11 @@ static int a6xx_gmu_start(struct a6xx_gmu *gmu)
/* Set the log wptr index
* note: downstream saves the value in poweroff and restores it here
*/
- gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_PWR_COL_CP_RESP, 0);
+ if (adreno_is_a7xx(adreno_gpu))
+ gmu_write(gmu, REG_A6XX_GMU_GENERAL_9, 0);
+ else
+ gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_PWR_COL_CP_RESP, 0);
+
gmu_write(gmu, REG_A6XX_GMU_CM3_SYSRESET, 0);
@@ -518,7 +523,9 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
if (IS_ERR(pdcptr))
goto err;
- if (adreno_is_a650(adreno_gpu) || adreno_is_a660_family(adreno_gpu))
+ if (adreno_is_a650(adreno_gpu) ||
+ adreno_is_a660_family(adreno_gpu) ||
+ adreno_is_a7xx(adreno_gpu))
pdc_in_aop = true;
else if (adreno_is_a618(adreno_gpu) || adreno_is_a640_family(adreno_gpu))
pdc_address_offset = 0x30090;
@@ -550,7 +557,8 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
gmu_write_rscc(gmu, REG_A6XX_RSCC_PDC_MATCH_VALUE_HI, 0x4514);
/* Load RSC sequencer uCode for sleep and wakeup */
- if (adreno_is_a650_family(adreno_gpu)) {
+ if (adreno_is_a650_family(adreno_gpu) ||
+ adreno_is_a7xx(adreno_gpu)) {
gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0, 0xeaaae5a0);
gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 1, 0xe1a1ebab);
gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 2, 0xa2e0a581);
@@ -635,11 +643,18 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
/* Set up the idle state for the GMU */
static void a6xx_gmu_power_config(struct a6xx_gmu *gmu)
{
+ struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
+ struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
+
/* Disable GMU WB/RB buffer */
gmu_write(gmu, REG_A6XX_GMU_SYS_BUS_CONFIG, 0x1);
gmu_write(gmu, REG_A6XX_GMU_ICACHE_CONFIG, 0x1);
gmu_write(gmu, REG_A6XX_GMU_DCACHE_CONFIG, 0x1);
+ /* A7xx knows better by default! */
+ if (adreno_is_a7xx(adreno_gpu))
+ return;
+
gmu_write(gmu, REG_A6XX_GMU_PWR_COL_INTER_FRAME_CTRL, 0x9c40400);
switch (gmu->idle_level) {
@@ -702,7 +717,7 @@ static int a6xx_gmu_fw_load(struct a6xx_gmu *gmu)
u32 itcm_base = 0x00000000;
u32 dtcm_base = 0x00040000;
- if (adreno_is_a650_family(adreno_gpu))
+ if (adreno_is_a650_family(adreno_gpu) || adreno_is_a7xx(adreno_gpu))
dtcm_base = 0x10004000;
if (gmu->legacy) {
@@ -751,14 +766,22 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
{
struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
+ u32 fence_range_lower, fence_range_upper;
int ret;
u32 chipid;
- if (adreno_is_a650_family(adreno_gpu)) {
+ /* Vote veto for FAL10 */
+ if (adreno_is_a650_family(adreno_gpu) || adreno_is_a7xx(adreno_gpu)) {
gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_CX_FALNEXT_INTF, 1);
gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_CX_FAL_INTF, 1);
}
+ /* Turn on TCM (Tightly Coupled Memory) retention */
+ if (adreno_is_a7xx(adreno_gpu))
+ a6xx_llc_write(a6xx_gpu, REG_A7XX_CX_MISC_TCM_RET_CNTL, 1);
+ else
+ gmu_write(gmu, REG_A6XX_GMU_GENERAL_7, 1);
+
if (state == GMU_WARM_BOOT) {
ret = a6xx_rpmh_start(gmu);
if (ret)
@@ -768,9 +791,6 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
"GMU firmware is not loaded\n"))
return -ENOENT;
- /* Turn on register retention */
- gmu_write(gmu, REG_A6XX_GMU_GENERAL_7, 1);
-
ret = a6xx_rpmh_start(gmu);
if (ret)
return ret;
@@ -780,6 +800,7 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
return ret;
}
+ /* Clear init result to make sure we are getting a fresh value */
gmu_write(gmu, REG_A6XX_GMU_CM3_FW_INIT_RESULT, 0);
gmu_write(gmu, REG_A6XX_GMU_CM3_BOOT_CONFIG, 0x02);
@@ -787,8 +808,18 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
gmu_write(gmu, REG_A6XX_GMU_HFI_QTBL_ADDR, gmu->hfi.iova);
gmu_write(gmu, REG_A6XX_GMU_HFI_QTBL_INFO, 1);
+ if (adreno_is_a7xx(adreno_gpu)) {
+ fence_range_upper = 0x32;
+ fence_range_lower = 0x8a0;
+ } else {
+ fence_range_upper = 0xa;
+ fence_range_lower = 0xa0;
+ }
+
gmu_write(gmu, REG_A6XX_GMU_AHB_FENCE_RANGE_0,
- (1 << 31) | (0xa << 18) | (0xa0));
+ BIT(31) |
+ FIELD_PREP(GENMASK(30, 18), fence_range_upper) |
+ FIELD_PREP(GENMASK(17, 0), fence_range_lower));
/*
* Snapshots toggle the NMI bit which will result in a jump to the NMI
@@ -807,10 +838,17 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
chipid |= (adreno_gpu->chip_id << 4) & 0xf000; /* minor */
chipid |= (adreno_gpu->chip_id << 8) & 0x0f00; /* patchid */
- gmu_write(gmu, REG_A6XX_GMU_HFI_SFR_ADDR, chipid);
+ if (adreno_is_a7xx(adreno_gpu)) {
+ gmu_write(gmu, REG_A6XX_GMU_GENERAL_10, chipid);
+ gmu_write(gmu, REG_A6XX_GMU_GENERAL_8,
+ (gmu->log.iova & GENMASK(31, 12)) |
+ ((gmu->log.size / SZ_4K - 1) & GENMASK(7, 0)));
+ } else {
+ gmu_write(gmu, REG_A6XX_GMU_HFI_SFR_ADDR, chipid);
- gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_PWR_COL_CP_MSG,
- gmu->log.iova | (gmu->log.size / SZ_4K - 1));
+ gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_PWR_COL_CP_MSG,
+ gmu->log.iova | (gmu->log.size / SZ_4K - 1));
+ }
/* Set up the lowest idle level on the GMU */
a6xx_gmu_power_config(gmu);
@@ -984,15 +1022,19 @@ int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
enable_irq(gmu->gmu_irq);
/* Check to see if we are doing a cold or warm boot */
- status = gmu_read(gmu, REG_A6XX_GMU_GENERAL_7) == 1 ?
- GMU_WARM_BOOT : GMU_COLD_BOOT;
-
- /*
- * Warm boot path does not work on newer GPUs
- * Presumably this is because icache/dcache regions must be restored
- */
- if (!gmu->legacy)
+ if (adreno_is_a7xx(adreno_gpu)) {
+ status = a6xx_llc_read(a6xx_gpu, REG_A7XX_CX_MISC_TCM_RET_CNTL) == 1 ?
+ GMU_WARM_BOOT : GMU_COLD_BOOT;
+ } else if (gmu->legacy) {
+ status = gmu_read(gmu, REG_A6XX_GMU_GENERAL_7) == 1 ?
+ GMU_WARM_BOOT : GMU_COLD_BOOT;
+ } else {
+ /*
+ * Warm boot path does not work on newer A6xx GPUs
+ * Presumably this is because icache/dcache regions must be restored
+ */
status = GMU_COLD_BOOT;
+ }
ret = a6xx_gmu_fw_start(gmu, status);
if (ret)
@@ -1604,7 +1646,8 @@ int a6xx_gmu_init(struct a6xx_gpu *a6xx_gpu, struct device_node *node)
* are otherwise unused by a660.
*/
gmu->dummy.size = SZ_4K;
- if (adreno_is_a660_family(adreno_gpu)) {
+ if (adreno_is_a660_family(adreno_gpu) ||
+ adreno_is_a7xx(adreno_gpu)) {
ret = a6xx_gmu_memory_alloc(gmu, &gmu->debug, SZ_4K * 7,
0x60400000, "debug");
if (ret)
@@ -1620,7 +1663,8 @@ int a6xx_gmu_init(struct a6xx_gpu *a6xx_gpu, struct device_node *node)
goto err_memory;
/* Note that a650 family also includes a660 family: */
- if (adreno_is_a650_family(adreno_gpu)) {
+ if (adreno_is_a650_family(adreno_gpu) ||
+ adreno_is_a7xx(adreno_gpu)) {
ret = a6xx_gmu_memory_alloc(gmu, &gmu->icache,
SZ_16M - SZ_16K, 0x04000, "icache");
if (ret)
@@ -1668,7 +1712,8 @@ int a6xx_gmu_init(struct a6xx_gpu *a6xx_gpu, struct device_node *node)
goto err_memory;
}
- if (adreno_is_a650_family(adreno_gpu)) {
+ if (adreno_is_a650_family(adreno_gpu) ||
+ adreno_is_a7xx(adreno_gpu)) {
gmu->rscc = a6xx_gmu_get_mmio(pdev, "rscc");
if (IS_ERR(gmu->rscc)) {
ret = -ENODEV;
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
index d4e85e24002f..61ce8d053355 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
@@ -103,6 +103,7 @@ static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
struct msm_ringbuffer *ring, struct msm_file_private *ctx)
{
bool sysprof = refcount_read(&a6xx_gpu->base.base.sysprof_active) > 1;
+ struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
phys_addr_t ttbr;
u32 asid;
u64 memptr = rbmemptr(ring, ttbr0);
@@ -114,9 +115,11 @@ static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
return;
if (!sysprof) {
- /* Turn off protected mode to write to special registers */
- OUT_PKT7(ring, CP_SET_PROTECTED_MODE, 1);
- OUT_RING(ring, 0);
+ if (!adreno_is_a7xx(adreno_gpu)) {
+ /* Turn off protected mode to write to special registers */
+ OUT_PKT7(ring, CP_SET_PROTECTED_MODE, 1);
+ OUT_RING(ring, 0);
+ }
OUT_PKT4(ring, REG_A6XX_RBBM_PERFCTR_SRAM_INIT_CMD, 1);
OUT_RING(ring, 1);
@@ -141,6 +144,16 @@ static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
OUT_RING(ring, lower_32_bits(ttbr));
OUT_RING(ring, (asid << 16) | upper_32_bits(ttbr));
+ /*
+ * Sync both threads after switching pagetables and enable BR only
+ * to make sure BV doesn't race ahead while BR is still switching
+ * pagetables.
+ */
+ if (adreno_is_a7xx(&a6xx_gpu->base)) {
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, CP_THREAD_CONTROL_0_SYNC_THREADS | CP_SET_THREAD_BR);
+ }
+
/*
* And finally, trigger a uche flush to be sure there isn't anything
* lingering in that part of the GPU
@@ -163,9 +176,11 @@ static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
OUT_RING(ring, CP_WAIT_REG_MEM_4_MASK(0x1));
OUT_RING(ring, CP_WAIT_REG_MEM_5_DELAY_LOOP_CYCLES(0));
- /* Re-enable protected mode: */
- OUT_PKT7(ring, CP_SET_PROTECTED_MODE, 1);
- OUT_RING(ring, 1);
+ if (!adreno_is_a7xx(adreno_gpu)) {
+ /* Re-enable protected mode: */
+ OUT_PKT7(ring, CP_SET_PROTECTED_MODE, 1);
+ OUT_RING(ring, 1);
+ }
}
}
@@ -252,6 +267,133 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
a6xx_flush(gpu, ring);
}
+static void a7xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
+{
+ unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
+ struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+ struct a6xx_gpu *a6xx_gpu = to_a6xx_gpu(adreno_gpu);
+ struct msm_ringbuffer *ring = submit->ring;
+ unsigned int i, ibs = 0;
+
+ /*
+ * Toggle concurrent binning for pagetable switch and set the thread to
+ * BR since only it can execute the pagetable switch packets.
+ */
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, CP_THREAD_CONTROL_0_SYNC_THREADS | CP_SET_THREAD_BR);
+
+ a6xx_set_pagetable(a6xx_gpu, ring, submit->queue->ctx);
+
+ get_stats_counter(ring, REG_A6XX_RBBM_PERFCTR_CP(0),
+ rbmemptr_stats(ring, index, cpcycles_start));
+ get_stats_counter(ring, REG_A6XX_CP_ALWAYS_ON_COUNTER,
+ rbmemptr_stats(ring, index, alwayson_start));
+
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, CP_SET_THREAD_BOTH);
+
+ OUT_PKT7(ring, CP_SET_MARKER, 1);
+ OUT_RING(ring, 0x101); /* IFPC disable */
+
+ OUT_PKT7(ring, CP_SET_MARKER, 1);
+ OUT_RING(ring, 0x00d); /* IB1LIST start */
+
+ /* Submit the commands */
+ for (i = 0; i < submit->nr_cmds; i++) {
+ switch (submit->cmd[i].type) {
+ case MSM_SUBMIT_CMD_IB_TARGET_BUF:
+ break;
+ case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
+ if (gpu->cur_ctx_seqno == submit->queue->ctx->seqno)
+ break;
+ fallthrough;
+ case MSM_SUBMIT_CMD_BUF:
+ OUT_PKT7(ring, CP_INDIRECT_BUFFER_PFE, 3);
+ OUT_RING(ring, lower_32_bits(submit->cmd[i].iova));
+ OUT_RING(ring, upper_32_bits(submit->cmd[i].iova));
+ OUT_RING(ring, submit->cmd[i].size);
+ ibs++;
+ break;
+ }
+
+ /*
+ * Periodically update shadow-wptr if needed, so that we
+ * can see partial progress of submits with large # of
+ * cmds.. otherwise we could needlessly stall waiting for
+ * ringbuffer state, simply due to looking at a shadow
+ * rptr value that has not been updated
+ */
+ if ((ibs % 32) == 0)
+ update_shadow_rptr(gpu, ring);
+ }
+
+ OUT_PKT7(ring, CP_SET_MARKER, 1);
+ OUT_RING(ring, 0x00e); /* IB1LIST end */
+
+ get_stats_counter(ring, REG_A6XX_RBBM_PERFCTR_CP(0),
+ rbmemptr_stats(ring, index, cpcycles_end));
+ get_stats_counter(ring, REG_A6XX_CP_ALWAYS_ON_COUNTER,
+ rbmemptr_stats(ring, index, alwayson_end));
+
+ /* Write the fence to the scratch register */
+ OUT_PKT4(ring, REG_A6XX_CP_SCRATCH_REG(2), 1);
+ OUT_RING(ring, submit->seqno);
+
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, CP_SET_THREAD_BR);
+
+ OUT_PKT7(ring, CP_EVENT_WRITE, 1);
+ OUT_RING(ring, CCU_INVALIDATE_DEPTH);
+
+ OUT_PKT7(ring, CP_EVENT_WRITE, 1);
+ OUT_RING(ring, CCU_INVALIDATE_COLOR);
+
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, CP_SET_THREAD_BV);
+
+ /*
+ * Make sure the timestamp is committed once BV pipe is
+ * completely done with this submission.
+ */
+ OUT_PKT7(ring, CP_EVENT_WRITE, 4);
+ OUT_RING(ring, CACHE_CLEAN | BIT(27));
+ OUT_RING(ring, lower_32_bits(rbmemptr(ring, bv_fence)));
+ OUT_RING(ring, upper_32_bits(rbmemptr(ring, bv_fence)));
+ OUT_RING(ring, submit->seqno);
+
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, CP_SET_THREAD_BR);
+
+ /*
+ * This makes sure that BR doesn't race ahead and commit
+ * timestamp to memstore while BV is still processing
+ * this submission.
+ */
+ OUT_PKT7(ring, CP_WAIT_TIMESTAMP, 4);
+ OUT_RING(ring, 0);
+ OUT_RING(ring, lower_32_bits(rbmemptr(ring, bv_fence)));
+ OUT_RING(ring, upper_32_bits(rbmemptr(ring, bv_fence)));
+ OUT_RING(ring, submit->seqno);
+
+ /* write the ringbuffer timestamp */
+ OUT_PKT7(ring, CP_EVENT_WRITE, 4);
+ OUT_RING(ring, CACHE_CLEAN | CP_EVENT_WRITE_0_IRQ | BIT(27));
+ OUT_RING(ring, lower_32_bits(rbmemptr(ring, fence)));
+ OUT_RING(ring, upper_32_bits(rbmemptr(ring, fence)));
+ OUT_RING(ring, submit->seqno);
+
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, CP_SET_THREAD_BOTH);
+
+ OUT_PKT7(ring, CP_SET_MARKER, 1);
+ OUT_RING(ring, 0x100); /* IFPC enable */
+
+ trace_msm_gpu_submit_flush(submit,
+ gpu_read64(gpu, REG_A6XX_CP_ALWAYS_ON_COUNTER));
+
+ a6xx_flush(gpu, ring);
+}
+
const struct adreno_reglist a612_hwcg[] = {
{REG_A6XX_RBBM_CLOCK_CNTL_SP0, 0x22222222},
{REG_A6XX_RBBM_CLOCK_CNTL2_SP0, 0x02222220},
@@ -714,6 +856,15 @@ static void a6xx_set_hwcg(struct msm_gpu *gpu, bool state)
else
clock_cntl_on = 0x8aa8aa82;
+ if (adreno_is_a7xx(adreno_gpu)) {
+ gmu_write(&a6xx_gpu->gmu, REG_A6XX_GPU_GMU_AO_GMU_CGC_MODE_CNTL,
+ state ? 0x20000 : 0);
+ gmu_write(&a6xx_gpu->gmu, REG_A6XX_GPU_GMU_AO_GMU_CGC_DELAY_CNTL,
+ state ? 0x10111 : 0);
+ gmu_write(&a6xx_gpu->gmu, REG_A6XX_GPU_GMU_AO_GMU_CGC_HYST_CNTL,
+ state ? 0x5555 : 0);
+ }
+
val = gpu_read(gpu, REG_A6XX_RBBM_CLOCK_CNTL);
/* Don't re-program the registers if they are already correct */
@@ -721,14 +872,14 @@ static void a6xx_set_hwcg(struct msm_gpu *gpu, bool state)
return;
/* Disable SP clock before programming HWCG registers */
- if (!adreno_is_a610(adreno_gpu))
+ if (!adreno_is_a610(adreno_gpu) && !adreno_is_a7xx(adreno_gpu))
gmu_rmw(gmu, REG_A6XX_GPU_GMU_GX_SPTPRAC_CLOCK_CONTROL, 1, 0);
for (i = 0; (reg = &adreno_gpu->info->hwcg[i], reg->offset); i++)
gpu_write(gpu, reg->offset, state ? reg->value : 0);
/* Enable SP clock */
- if (!adreno_is_a610(adreno_gpu))
+ if (!adreno_is_a610(adreno_gpu) && !adreno_is_a7xx(adreno_gpu))
gmu_rmw(gmu, REG_A6XX_GPU_GMU_GX_SPTPRAC_CLOCK_CONTROL, 0, 1);
gpu_write(gpu, REG_A6XX_RBBM_CLOCK_CNTL, state ? clock_cntl_on : 0);
@@ -1017,6 +1168,10 @@ static void a6xx_set_ubwc_config(struct msm_gpu *gpu)
uavflagprd_inv << 4 | min_acc_len << 3 |
hbb_lo << 1 | ubwc_mode);
+ if (adreno_is_a7xx(adreno_gpu))
+ gpu_write(gpu, REG_A7XX_GRAS_NC_MODE_CNTL,
+ FIELD_PREP(GENMASK(8, 5), hbb_lo));
+
gpu_write(gpu, REG_A6XX_UCHE_MODE_CNTL, min_acc_len << 23 | hbb_lo << 21);
}
@@ -1049,6 +1204,55 @@ static int a6xx_cp_init(struct msm_gpu *gpu)
return a6xx_idle(gpu, ring) ? 0 : -EINVAL;
}
+static int a7xx_cp_init(struct msm_gpu *gpu)
+{
+ struct msm_ringbuffer *ring = gpu->rb[0];
+ u32 mask;
+
+ /* Disable concurrent binning before sending CP init */
+ OUT_PKT7(ring, CP_THREAD_CONTROL, 1);
+ OUT_RING(ring, BIT(27));
+
+ OUT_PKT7(ring, CP_ME_INIT, 7);
+
+ /* Use multiple HW contexts */
+ mask = BIT(0);
+
+ /* Enable error detection */
+ mask |= BIT(1);
+
+ /* Set default reset state */
+ mask |= BIT(3);
+
+ /* Disable save/restore of performance counters across preemption */
+ mask |= BIT(6);
+
+ /* Enable the register init list with the spinlock */
+ mask |= BIT(8);
+
+ OUT_RING(ring, mask);
+
+ /* Enable multiple hardware contexts */
+ OUT_RING(ring, 0x00000003);
+
+ /* Enable error detection */
+ OUT_RING(ring, 0x20000000);
+
+ /* Operation mode mask */
+ OUT_RING(ring, 0x00000002);
+
+ /* *Don't* send a power up reg list for concurrent binning (TODO) */
+ /* Lo address */
+ OUT_RING(ring, 0x00000000);
+ /* Hi address */
+ OUT_RING(ring, 0x00000000);
+ /* BIT(31) set => read the regs from the list */
+ OUT_RING(ring, 0x00000000);
+
+ a6xx_flush(gpu, ring);
+ return a6xx_idle(gpu, ring) ? 0 : -EINVAL;
+}
+
/*
* Check that the microcode version is new enough to include several key
* security fixes. Return true if the ucode is safe.
@@ -1065,6 +1269,10 @@ static bool a6xx_ucode_check_version(struct a6xx_gpu *a6xx_gpu,
if (IS_ERR(buf))
return false;
+ /* A7xx is safe! */
+ if (adreno_is_a7xx(adreno_gpu))
+ return true;
+
/*
* Targets up to a640 (a618, a630 and a640) need to check for a
* microcode version that is patched to support the whereami opcode or
@@ -1181,16 +1389,39 @@ static int a6xx_zap_shader_init(struct msm_gpu *gpu)
}
#define A6XX_INT_MASK (A6XX_RBBM_INT_0_MASK_CP_AHB_ERROR | \
- A6XX_RBBM_INT_0_MASK_RBBM_ATB_ASYNCFIFO_OVERFLOW | \
- A6XX_RBBM_INT_0_MASK_CP_HW_ERROR | \
- A6XX_RBBM_INT_0_MASK_CP_IB2 | \
- A6XX_RBBM_INT_0_MASK_CP_IB1 | \
- A6XX_RBBM_INT_0_MASK_CP_RB | \
- A6XX_RBBM_INT_0_MASK_CP_CACHE_FLUSH_TS | \
- A6XX_RBBM_INT_0_MASK_RBBM_ATB_BUS_OVERFLOW | \
- A6XX_RBBM_INT_0_MASK_RBBM_HANG_DETECT | \
- A6XX_RBBM_INT_0_MASK_UCHE_OOB_ACCESS | \
- A6XX_RBBM_INT_0_MASK_UCHE_TRAP_INTR)
+ A6XX_RBBM_INT_0_MASK_RBBM_ATB_ASYNCFIFO_OVERFLOW | \
+ A6XX_RBBM_INT_0_MASK_CP_HW_ERROR | \
+ A6XX_RBBM_INT_0_MASK_CP_IB2 | \
+ A6XX_RBBM_INT_0_MASK_CP_IB1 | \
+ A6XX_RBBM_INT_0_MASK_CP_RB | \
+ A6XX_RBBM_INT_0_MASK_CP_CACHE_FLUSH_TS | \
+ A6XX_RBBM_INT_0_MASK_RBBM_ATB_BUS_OVERFLOW | \
+ A6XX_RBBM_INT_0_MASK_RBBM_HANG_DETECT | \
+ A6XX_RBBM_INT_0_MASK_UCHE_OOB_ACCESS | \
+ A6XX_RBBM_INT_0_MASK_UCHE_TRAP_INTR)
+
+#define A7XX_INT_MASK (A6XX_RBBM_INT_0_MASK_CP_AHB_ERROR | \
+ A6XX_RBBM_INT_0_MASK_RBBM_ATB_ASYNCFIFO_OVERFLOW | \
+ A6XX_RBBM_INT_0_MASK_RBBM_GPC_ERROR | \
+ A6XX_RBBM_INT_0_MASK_CP_SW | \
+ A6XX_RBBM_INT_0_MASK_CP_HW_ERROR | \
+ A6XX_RBBM_INT_0_MASK_PM4CPINTERRUPT | \
+ A6XX_RBBM_INT_0_MASK_CP_RB_DONE_TS | \
+ A6XX_RBBM_INT_0_MASK_CP_CACHE_FLUSH_TS | \
+ A6XX_RBBM_INT_0_MASK_RBBM_ATB_BUS_OVERFLOW | \
+ A6XX_RBBM_INT_0_MASK_RBBM_HANG_DETECT | \
+ A6XX_RBBM_INT_0_MASK_UCHE_OOB_ACCESS | \
+ A6XX_RBBM_INT_0_MASK_UCHE_TRAP_INTR | \
+ A6XX_RBBM_INT_0_MASK_TSBWRITEERROR)
+
+#define A7XX_APRIV_MASK (A6XX_CP_APRIV_CNTL_ICACHE | \
+ A6XX_CP_APRIV_CNTL_RBFETCH | \
+ A6XX_CP_APRIV_CNTL_RBPRIVLEVEL | \
+ A6XX_CP_APRIV_CNTL_RBRPWB)
+
+#define A7XX_BR_APRIVMASK (A7XX_APRIV_MASK | \
+ A6XX_CP_APRIV_CNTL_CDREAD | \
+ A6XX_CP_APRIV_CNTL_CDWRITE)
static int hw_init(struct msm_gpu *gpu)
{
@@ -1232,19 +1463,21 @@ static int hw_init(struct msm_gpu *gpu)
gpu_write64(gpu, REG_A6XX_RBBM_SECVID_TSB_TRUSTED_BASE, 0x00000000);
gpu_write(gpu, REG_A6XX_RBBM_SECVID_TSB_TRUSTED_SIZE, 0x00000000);
- /* Turn on 64 bit addressing for all blocks */
- gpu_write(gpu, REG_A6XX_CP_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_VSC_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_GRAS_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_RB_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_PC_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_HLSQ_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_VFD_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_VPC_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_UCHE_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_SP_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_TPL1_ADDR_MODE_CNTL, 0x1);
- gpu_write(gpu, REG_A6XX_RBBM_SECVID_TSB_ADDR_MODE_CNTL, 0x1);
+ if (!adreno_is_a7xx(adreno_gpu)) {
+ /* Turn on 64 bit addressing for all blocks */
+ gpu_write(gpu, REG_A6XX_CP_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_VSC_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_GRAS_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_RB_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_PC_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_HLSQ_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_VFD_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_VPC_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_UCHE_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_SP_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_TPL1_ADDR_MODE_CNTL, 0x1);
+ gpu_write(gpu, REG_A6XX_RBBM_SECVID_TSB_ADDR_MODE_CNTL, 0x1);
+ }
/* enable hardware clockgating */
a6xx_set_hwcg(gpu, true);
@@ -1252,12 +1485,14 @@ static int hw_init(struct msm_gpu *gpu)
/* VBIF/GBIF start*/
if (adreno_is_a610(adreno_gpu) ||
adreno_is_a640_family(adreno_gpu) ||
- adreno_is_a650_family(adreno_gpu)) {
+ adreno_is_a650_family(adreno_gpu) ||
+ adreno_is_a7xx(adreno_gpu)) {
gpu_write(gpu, REG_A6XX_GBIF_QSB_SIDE0, 0x00071620);
gpu_write(gpu, REG_A6XX_GBIF_QSB_SIDE1, 0x00071620);
gpu_write(gpu, REG_A6XX_GBIF_QSB_SIDE2, 0x00071620);
gpu_write(gpu, REG_A6XX_GBIF_QSB_SIDE3, 0x00071620);
- gpu_write(gpu, REG_A6XX_RBBM_GBIF_CLIENT_QOS_CNTL, 0x3);
+ gpu_write(gpu, REG_A6XX_RBBM_GBIF_CLIENT_QOS_CNTL,
+ adreno_is_a7xx(adreno_gpu) ? 0x2120212 : 0x3);
} else {
gpu_write(gpu, REG_A6XX_RBBM_VBIF_CLIENT_QOS_CNTL, 0x3);
}
@@ -1265,13 +1500,21 @@ static int hw_init(struct msm_gpu *gpu)
if (adreno_is_a630(adreno_gpu))
gpu_write(gpu, REG_A6XX_VBIF_GATE_OFF_WRREQ_EN, 0x00000009);
+ if (adreno_is_a7xx(adreno_gpu))
+ gpu_write(gpu, REG_A6XX_UCHE_GBIF_GX_CONFIG, 0x10240e0);
+
/* Make all blocks contribute to the GPU BUSY perf counter */
gpu_write(gpu, REG_A6XX_RBBM_PERFCTR_GPU_BUSY_MASKED, 0xffffffff);
/* Disable L2 bypass in the UCHE */
- gpu_write64(gpu, REG_A6XX_UCHE_WRITE_RANGE_MAX, 0x0001ffffffffffc0llu);
- gpu_write64(gpu, REG_A6XX_UCHE_TRAP_BASE, 0x0001fffffffff000llu);
- gpu_write64(gpu, REG_A6XX_UCHE_WRITE_THRU_BASE, 0x0001fffffffff000llu);
+ if (adreno_is_a7xx(adreno_gpu)) {
+ gpu_write64(gpu, REG_A6XX_UCHE_TRAP_BASE, 0x0001fffffffff000llu);
+ gpu_write64(gpu, REG_A6XX_UCHE_WRITE_THRU_BASE, 0x0001fffffffff000llu);
+ } else {
+ gpu_write64(gpu, REG_A6XX_UCHE_WRITE_RANGE_MAX, 0x0001ffffffffffc0llu);
+ gpu_write64(gpu, REG_A6XX_UCHE_TRAP_BASE, 0x0001fffffffff000llu);
+ gpu_write64(gpu, REG_A6XX_UCHE_WRITE_THRU_BASE, 0x0001fffffffff000llu);
+ }
if (!adreno_is_a650_family(adreno_gpu)) {
/* Set the GMEM VA range [0x100000:0x100000 + gpu->gmem - 1] */
@@ -1281,8 +1524,12 @@ static int hw_init(struct msm_gpu *gpu)
0x00100000 + adreno_gpu->info->gmem - 1);
}
- gpu_write(gpu, REG_A6XX_UCHE_FILTER_CNTL, 0x804);
- gpu_write(gpu, REG_A6XX_UCHE_CACHE_WAYS, 0x4);
+ if (adreno_is_a7xx(adreno_gpu))
+ gpu_write(gpu, REG_A6XX_UCHE_CACHE_WAYS, BIT(23));
+ else {
+ gpu_write(gpu, REG_A6XX_UCHE_FILTER_CNTL, 0x804);
+ gpu_write(gpu, REG_A6XX_UCHE_CACHE_WAYS, 0x4);
+ }
if (adreno_is_a640_family(adreno_gpu) || adreno_is_a650_family(adreno_gpu)) {
gpu_write(gpu, REG_A6XX_CP_ROQ_THRESHOLDS_2, 0x02000140);
@@ -1290,7 +1537,7 @@ static int hw_init(struct msm_gpu *gpu)
} else if (adreno_is_a610(adreno_gpu)) {
gpu_write(gpu, REG_A6XX_CP_ROQ_THRESHOLDS_2, 0x00800060);
gpu_write(gpu, REG_A6XX_CP_ROQ_THRESHOLDS_1, 0x40201b16);
- } else {
+ } else if (!adreno_is_a7xx(adreno_gpu)) {
gpu_write(gpu, REG_A6XX_CP_ROQ_THRESHOLDS_2, 0x010000c0);
gpu_write(gpu, REG_A6XX_CP_ROQ_THRESHOLDS_1, 0x8040362c);
}
@@ -1302,7 +1549,7 @@ static int hw_init(struct msm_gpu *gpu)
if (adreno_is_a610(adreno_gpu)) {
gpu_write(gpu, REG_A6XX_CP_MEM_POOL_SIZE, 48);
gpu_write(gpu, REG_A6XX_CP_MEM_POOL_DBG_ADDR, 47);
- } else
+ } else if (!adreno_is_a7xx(adreno_gpu))
gpu_write(gpu, REG_A6XX_CP_MEM_POOL_SIZE, 128);
/* Setting the primFifo thresholds default values,
@@ -1318,7 +1565,7 @@ static int hw_init(struct msm_gpu *gpu)
gpu_write(gpu, REG_A6XX_PC_DBG_ECO_CNTL, 0x00018000);
else if (adreno_is_a610(adreno_gpu))
gpu_write(gpu, REG_A6XX_PC_DBG_ECO_CNTL, 0x00080000);
- else
+ else if (!adreno_is_a7xx(adreno_gpu))
gpu_write(gpu, REG_A6XX_PC_DBG_ECO_CNTL, 0x00180000);
/* Set the AHB default slave response to "ERROR" */
@@ -1327,6 +1574,12 @@ static int hw_init(struct msm_gpu *gpu)
/* Turn on performance counters */
gpu_write(gpu, REG_A6XX_RBBM_PERFCTR_CNTL, 0x1);
+ if (adreno_is_a7xx(adreno_gpu)) {
+ /* Turn on the IFPC counter (countable 4 on XOCLK4) */
+ gmu_write(&a6xx_gpu->gmu, REG_A6XX_GMU_CX_GMU_POWER_COUNTER_SELECT_1,
+ FIELD_PREP(GENMASK(7, 0), 0x4));
+ }
+
/* Select CP0 to always count cycles */
gpu_write(gpu, REG_A6XX_CP_PERFCTR_CP_SEL(0), PERF_CP_ALWAYS_COUNT);
@@ -1373,15 +1626,31 @@ static int hw_init(struct msm_gpu *gpu)
/* Set dualQ + disable afull for A660 GPU */
if (adreno_is_a660(adreno_gpu))
gpu_write(gpu, REG_A6XX_UCHE_CMDQ_CONFIG, 0x66906);
+ else if (adreno_is_a7xx(adreno_gpu))
+ gpu_write(gpu, REG_A6XX_UCHE_CMDQ_CONFIG,
+ FIELD_PREP(GENMASK(19, 16), 6) |
+ FIELD_PREP(GENMASK(15, 12), 6) |
+ FIELD_PREP(GENMASK(11, 8), 9) |
+ BIT(3) | BIT(2) |
+ FIELD_PREP(GENMASK(1, 0), 2));
/* Enable expanded apriv for targets that support it */
if (gpu->hw_apriv) {
- gpu_write(gpu, REG_A6XX_CP_APRIV_CNTL,
- (1 << 6) | (1 << 5) | (1 << 3) | (1 << 2) | (1 << 1));
+ if (adreno_is_a7xx(adreno_gpu)) {
+ gpu_write(gpu, REG_A6XX_CP_APRIV_CNTL,
+ A7XX_BR_APRIVMASK);
+ gpu_write(gpu, REG_A7XX_CP_BV_APRIV_CNTL,
+ A7XX_APRIV_MASK);
+ gpu_write(gpu, REG_A7XX_CP_LPAC_APRIV_CNTL,
+ A7XX_APRIV_MASK);
+ } else
+ gpu_write(gpu, REG_A6XX_CP_APRIV_CNTL,
+ BIT(6) | BIT(5) | BIT(3) | BIT(2) | BIT(1));
}
/* Enable interrupts */
- gpu_write(gpu, REG_A6XX_RBBM_INT_0_MASK, A6XX_INT_MASK);
+ gpu_write(gpu, REG_A6XX_RBBM_INT_0_MASK,
+ adreno_is_a7xx(adreno_gpu) ? A7XX_INT_MASK : A6XX_INT_MASK);
ret = adreno_hw_init(gpu);
if (ret)
@@ -1408,6 +1677,12 @@ static int hw_init(struct msm_gpu *gpu)
shadowptr(a6xx_gpu, gpu->rb[0]));
}
+ /* ..which means "always" on A7xx, also for BV shadow */
+ if (adreno_is_a7xx(adreno_gpu)) {
+ gpu_write64(gpu, REG_A7XX_CP_BV_RB_RPTR_ADDR,
+ rbmemptr(gpu->rb[0], bv_fence));
+ }
+
/* Always come up on rb 0 */
a6xx_gpu->cur_ring = gpu->rb[0];
@@ -1416,7 +1691,7 @@ static int hw_init(struct msm_gpu *gpu)
/* Enable the SQE_to start the CP engine */
gpu_write(gpu, REG_A6XX_CP_SQE_CNTL, 1);
- ret = a6xx_cp_init(gpu);
+ ret = adreno_is_a7xx(adreno_gpu) ? a7xx_cp_init(gpu) : a6xx_cp_init(gpu);
if (ret)
goto out;
@@ -1653,7 +1928,7 @@ static void a6xx_cp_hw_err_irq(struct msm_gpu *gpu)
(val & 0x3ffff), val);
}
- if (status & A6XX_CP_INT_CP_AHB_ERROR)
+ if (status & A6XX_CP_INT_CP_AHB_ERROR && !adreno_is_a7xx(to_adreno_gpu(gpu)))
dev_err_ratelimited(&gpu->pdev->dev, "CP AHB error interrupt\n");
if (status & A6XX_CP_INT_CP_VSD_PARITY_ERROR)
@@ -1803,6 +2078,35 @@ static void a6xx_llc_activate(struct a6xx_gpu *a6xx_gpu)
gpu_rmw(gpu, REG_A6XX_GBIF_SCACHE_CNTL1, GENMASK(24, 0), cntl1_regval);
}
+static void a7xx_llc_activate(struct a6xx_gpu *a6xx_gpu)
+{
+ struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
+ struct msm_gpu *gpu = &adreno_gpu->base;
+
+ if (IS_ERR(a6xx_gpu->llc_mmio))
+ return;
+
+ if (!llcc_slice_activate(a6xx_gpu->llc_slice)) {
+ u32 gpu_scid = llcc_get_slice_id(a6xx_gpu->llc_slice);
+
+ gpu_scid &= GENMASK(4, 0);
+
+ gpu_write(gpu, REG_A6XX_GBIF_SCACHE_CNTL1,
+ FIELD_PREP(GENMASK(29, 25), gpu_scid) |
+ FIELD_PREP(GENMASK(24, 20), gpu_scid) |
+ FIELD_PREP(GENMASK(19, 15), gpu_scid) |
+ FIELD_PREP(GENMASK(14, 10), gpu_scid) |
+ FIELD_PREP(GENMASK(9, 5), gpu_scid) |
+ FIELD_PREP(GENMASK(4, 0), gpu_scid));
+
+ gpu_write(gpu, REG_A6XX_GBIF_SCACHE_CNTL0,
+ FIELD_PREP(GENMASK(14, 10), gpu_scid) |
+ BIT(8));
+ }
+
+ llcc_slice_activate(a6xx_gpu->htw_llc_slice);
+}
+
static void a6xx_llc_slices_destroy(struct a6xx_gpu *a6xx_gpu)
{
/* No LLCC on non-RPMh (and by extension, non-GMU) SoCs */
@@ -1814,7 +2118,7 @@ static void a6xx_llc_slices_destroy(struct a6xx_gpu *a6xx_gpu)
}
static void a6xx_llc_slices_init(struct platform_device *pdev,
- struct a6xx_gpu *a6xx_gpu)
+ struct a6xx_gpu *a6xx_gpu, bool is_a7xx)
{
struct device_node *phandle;
@@ -1823,18 +2127,18 @@ static void a6xx_llc_slices_init(struct platform_device *pdev,
return;
/*
- * There is a different programming path for targets with an mmu500
- * attached, so detect if that is the case
+ * There is a different programming path for A6xx targets with an
+ * mmu500 attached, so detect if that is the case
*/
phandle = of_parse_phandle(pdev->dev.of_node, "iommus", 0);
a6xx_gpu->have_mmu500 = (phandle &&
of_device_is_compatible(phandle, "arm,mmu-500"));
of_node_put(phandle);
- if (a6xx_gpu->have_mmu500)
- a6xx_gpu->llc_mmio = NULL;
- else
+ if (is_a7xx || !a6xx_gpu->have_mmu500)
a6xx_gpu->llc_mmio = msm_ioremap(pdev, "cx_mem");
+ else
+ a6xx_gpu->llc_mmio = NULL;
a6xx_gpu->llc_slice = llcc_slice_getd(LLCC_GPU);
a6xx_gpu->htw_llc_slice = llcc_slice_getd(LLCC_GPUHTW);
@@ -1920,7 +2224,7 @@ static int a6xx_gmu_pm_resume(struct msm_gpu *gpu)
msm_devfreq_resume(gpu);
- a6xx_llc_activate(a6xx_gpu);
+ adreno_is_a7xx(adreno_gpu) ? a7xx_llc_activate : a6xx_llc_activate(a6xx_gpu);
return ret;
}
@@ -2307,6 +2611,37 @@ static const struct adreno_gpu_funcs funcs_gmuwrapper = {
.get_timestamp = a6xx_get_timestamp,
};
+static const struct adreno_gpu_funcs funcs_a7xx = {
+ .base = {
+ .get_param = adreno_get_param,
+ .set_param = adreno_set_param,
+ .hw_init = a6xx_hw_init,
+ .ucode_load = a6xx_ucode_load,
+ .pm_suspend = a6xx_gmu_pm_suspend,
+ .pm_resume = a6xx_gmu_pm_resume,
+ .recover = a6xx_recover,
+ .submit = a7xx_submit,
+ .active_ring = a6xx_active_ring,
+ .irq = a6xx_irq,
+ .destroy = a6xx_destroy,
+#if defined(CONFIG_DRM_MSM_GPU_STATE)
+ .show = a6xx_show,
+#endif
+ .gpu_busy = a6xx_gpu_busy,
+ .gpu_get_freq = a6xx_gmu_get_freq,
+ .gpu_set_freq = a6xx_gpu_set_freq,
+#if defined(CONFIG_DRM_MSM_GPU_STATE)
+ .gpu_state_get = a6xx_gpu_state_get,
+ .gpu_state_put = a6xx_gpu_state_put,
+#endif
+ .create_address_space = a6xx_create_address_space,
+ .create_private_address_space = a6xx_create_private_address_space,
+ .get_rptr = a6xx_get_rptr,
+ .progress = a6xx_progress,
+ },
+ .get_timestamp = a6xx_gmu_get_timestamp,
+};
+
struct msm_gpu *a6xx_gpu_init(struct drm_device *dev)
{
struct msm_drm_private *priv = dev->dev_private;
@@ -2316,6 +2651,7 @@ struct msm_gpu *a6xx_gpu_init(struct drm_device *dev)
struct a6xx_gpu *a6xx_gpu;
struct adreno_gpu *adreno_gpu;
struct msm_gpu *gpu;
+ bool is_a7xx;
int ret;
a6xx_gpu = kzalloc(sizeof(*a6xx_gpu), GFP_KERNEL);
@@ -2339,7 +2675,10 @@ struct msm_gpu *a6xx_gpu_init(struct drm_device *dev)
adreno_gpu->base.hw_apriv =
!!(config->info->quirks & ADRENO_QUIRK_HAS_HW_APRIV);
- a6xx_llc_slices_init(pdev, a6xx_gpu);
+ /* gpu->info only gets assigned in adreno_gpu_init() */
+ is_a7xx = config->info->family == ADRENO_7XX_GEN1;
+
+ a6xx_llc_slices_init(pdev, a6xx_gpu, is_a7xx);
ret = a6xx_set_supported_hw(&pdev->dev, config->info);
if (ret) {
@@ -2347,7 +2686,9 @@ struct msm_gpu *a6xx_gpu_init(struct drm_device *dev)
return ERR_PTR(ret);
}
- if (adreno_has_gmu_wrapper(adreno_gpu))
+ if (is_a7xx)
+ ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs_a7xx, 1);
+ else if (adreno_has_gmu_wrapper(adreno_gpu))
ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs_gmuwrapper, 1);
else
ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs, 1);
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index 8090dde03280..ea59724f8e41 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -567,6 +567,7 @@ int adreno_hw_init(struct msm_gpu *gpu)
ring->cur = ring->start;
ring->next = ring->start;
ring->memptrs->rptr = 0;
+ ring->memptrs->bv_fence = ring->fctx->completed_fence;
/* Detect and clean up an impossible fence, ie. if GPU managed
* to scribble something invalid, we don't want that to confuse
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index 49f38edf9854..ad07e788a69c 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -46,6 +46,7 @@ enum adreno_family {
ADRENO_6XX_GEN2, /* a640 family */
ADRENO_6XX_GEN3, /* a650 family */
ADRENO_6XX_GEN4, /* a660 family */
+ ADRENO_7XX_GEN1, /* a730 family */
};
#define ADRENO_QUIRK_TWO_PASS_USE_WFI BIT(0)
@@ -391,7 +392,8 @@ static inline int adreno_is_a650_family(const struct adreno_gpu *gpu)
{
if (WARN_ON_ONCE(!gpu->info))
return false;
- return gpu->info->family >= ADRENO_6XX_GEN3;
+ return gpu->info->family == ADRENO_6XX_GEN3 ||
+ adreno_is_a660_family(gpu);
}
static inline int adreno_is_a640_family(const struct adreno_gpu *gpu)
@@ -401,6 +403,12 @@ static inline int adreno_is_a640_family(const struct adreno_gpu *gpu)
return gpu->info->family == ADRENO_6XX_GEN2;
}
+static inline int adreno_is_a7xx(struct adreno_gpu *gpu)
+{
+ /* Update with non-fake (i.e. non-A702) Gen 7 GPUs */
+ return gpu->info->family == ADRENO_7XX_GEN1;
+}
+
u64 adreno_private_address_space_size(struct msm_gpu *gpu);
int adreno_get_param(struct msm_gpu *gpu, struct msm_file_private *ctx,
uint32_t param, uint64_t *value, uint32_t *len);
diff --git a/drivers/gpu/drm/msm/msm_ringbuffer.h b/drivers/gpu/drm/msm/msm_ringbuffer.h
index 698b333abccd..0d6beb8cd39a 100644
--- a/drivers/gpu/drm/msm/msm_ringbuffer.h
+++ b/drivers/gpu/drm/msm/msm_ringbuffer.h
@@ -30,6 +30,8 @@ struct msm_gpu_submit_stats {
struct msm_rbmemptrs {
volatile uint32_t rptr;
volatile uint32_t fence;
+ /* Introduced on A7xx */
+ volatile uint32_t bv_fence;
volatile struct msm_gpu_submit_stats stats[MSM_GPU_SUBMIT_STATS_COUNT];
volatile u64 ttbr0;
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* Re: [PATCH v3 05/10] drm/msm/a6xx: Add skeleton A7xx support
2023-08-23 12:55 ` [PATCH v3 05/10] drm/msm/a6xx: Add skeleton A7xx support Konrad Dybcio
@ 2023-09-11 15:30 ` Dmitry Baryshkov
0 siblings, 0 replies; 13+ messages in thread
From: Dmitry Baryshkov @ 2023-09-11 15:30 UTC (permalink / raw)
To: Konrad Dybcio, Rob Clark, Abhinav Kumar, Sean Paul, David Airlie,
Daniel Vetter, Rob Herring, Krzysztof Kozlowski, Conor Dooley,
Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Neil Armstrong
On 23/08/2023 15:55, Konrad Dybcio wrote:
> A7xx GPUs are - from kernel's POV anyway - basically another generation
> of A6xx. They build upon the A650/A660_family advancements, skipping some
> writes (presumably more values are preset correctly on reset), adding
> some new ones and changing others.
>
> One notable difference is the introduction of a second shadow, called BV.
> To handle this with the current code, allocate it right after the current
> RPTR shadow.
>
> BV handling and .submit are mostly based on Jonathan Marek's work.
>
> All A7xx GPUs are assumed to have a GMU.
> A702 is not an A7xx-class GPU, it's a weird forked A610.
>
> Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
> Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
> ---
> drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 95 +++++--
> drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 451 ++++++++++++++++++++++++++++----
> drivers/gpu/drm/msm/adreno/adreno_gpu.c | 1 +
> drivers/gpu/drm/msm/adreno/adreno_gpu.h | 10 +-
> drivers/gpu/drm/msm/msm_ringbuffer.h | 2 +
> 5 files changed, 478 insertions(+), 81 deletions(-)
>
> diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
> index 03fa89bf3e4b..75984260898e 100644
> --- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
> +++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
> @@ -200,9 +200,10 @@ int a6xx_gmu_wait_for_idle(struct a6xx_gmu *gmu)
>
> static int a6xx_gmu_start(struct a6xx_gmu *gmu)
> {
> + struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
> + struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
> + u32 mask, reset_val, val;
> int ret;
> - u32 val;
> - u32 mask, reset_val;
>
> val = gmu_read(gmu, REG_A6XX_GMU_CM3_DTCM_START + 0xff8);
> if (val <= 0x20010004) {
> @@ -218,7 +219,11 @@ static int a6xx_gmu_start(struct a6xx_gmu *gmu)
> /* Set the log wptr index
> * note: downstream saves the value in poweroff and restores it here
> */
> - gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_PWR_COL_CP_RESP, 0);
> + if (adreno_is_a7xx(adreno_gpu))
> + gmu_write(gmu, REG_A6XX_GMU_GENERAL_9, 0);
> + else
> + gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_PWR_COL_CP_RESP, 0);
> +
>
> gmu_write(gmu, REG_A6XX_GMU_CM3_SYSRESET, 0);
>
> @@ -518,7 +523,9 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
> if (IS_ERR(pdcptr))
> goto err;
>
> - if (adreno_is_a650(adreno_gpu) || adreno_is_a660_family(adreno_gpu))
> + if (adreno_is_a650(adreno_gpu) ||
> + adreno_is_a660_family(adreno_gpu) ||
> + adreno_is_a7xx(adreno_gpu))
> pdc_in_aop = true;
> else if (adreno_is_a618(adreno_gpu) || adreno_is_a640_family(adreno_gpu))
> pdc_address_offset = 0x30090;
> @@ -550,7 +557,8 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
> gmu_write_rscc(gmu, REG_A6XX_RSCC_PDC_MATCH_VALUE_HI, 0x4514);
>
> /* Load RSC sequencer uCode for sleep and wakeup */
> - if (adreno_is_a650_family(adreno_gpu)) {
> + if (adreno_is_a650_family(adreno_gpu) ||
> + adreno_is_a7xx(adreno_gpu)) {
> gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0, 0xeaaae5a0);
> gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 1, 0xe1a1ebab);
> gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 2, 0xa2e0a581);
> @@ -635,11 +643,18 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
> /* Set up the idle state for the GMU */
> static void a6xx_gmu_power_config(struct a6xx_gmu *gmu)
> {
> + struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
> + struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
> +
> /* Disable GMU WB/RB buffer */
> gmu_write(gmu, REG_A6XX_GMU_SYS_BUS_CONFIG, 0x1);
> gmu_write(gmu, REG_A6XX_GMU_ICACHE_CONFIG, 0x1);
> gmu_write(gmu, REG_A6XX_GMU_DCACHE_CONFIG, 0x1);
>
> + /* A7xx knows better by default! */
> + if (adreno_is_a7xx(adreno_gpu))
> + return;
> +
> gmu_write(gmu, REG_A6XX_GMU_PWR_COL_INTER_FRAME_CTRL, 0x9c40400);
>
> switch (gmu->idle_level) {
> @@ -702,7 +717,7 @@ static int a6xx_gmu_fw_load(struct a6xx_gmu *gmu)
> u32 itcm_base = 0x00000000;
> u32 dtcm_base = 0x00040000;
>
> - if (adreno_is_a650_family(adreno_gpu))
> + if (adreno_is_a650_family(adreno_gpu) || adreno_is_a7xx(adreno_gpu))
> dtcm_base = 0x10004000;
>
> if (gmu->legacy) {
> @@ -751,14 +766,22 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
> {
> struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
> struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
> + u32 fence_range_lower, fence_range_upper;
> int ret;
> u32 chipid;
>
> - if (adreno_is_a650_family(adreno_gpu)) {
> + /* Vote veto for FAL10 */
> + if (adreno_is_a650_family(adreno_gpu) || adreno_is_a7xx(adreno_gpu)) {
> gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_CX_FALNEXT_INTF, 1);
> gmu_write(gmu, REG_A6XX_GPU_GMU_CX_GMU_CX_FAL_INTF, 1);
> }
>
> + /* Turn on TCM (Tightly Coupled Memory) retention */
> + if (adreno_is_a7xx(adreno_gpu))
> + a6xx_llc_write(a6xx_gpu, REG_A7XX_CX_MISC_TCM_RET_CNTL, 1);
> + else
> + gmu_write(gmu, REG_A6XX_GMU_GENERAL_7, 1);
> +
> if (state == GMU_WARM_BOOT) {
> ret = a6xx_rpmh_start(gmu);
> if (ret)
> @@ -768,9 +791,6 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
> "GMU firmware is not loaded\n"))
> return -ENOENT;
>
> - /* Turn on register retention */
> - gmu_write(gmu, REG_A6XX_GMU_GENERAL_7, 1);
> -
> ret = a6xx_rpmh_start(gmu);
> if (ret)
> return ret;
> @@ -780,6 +800,7 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
> return ret;
> }
>
> + /* Clear init result to make sure we are getting a fresh value */
> gmu_write(gmu, REG_A6XX_GMU_CM3_FW_INIT_RESULT, 0);
> gmu_write(gmu, REG_A6XX_GMU_CM3_BOOT_CONFIG, 0x02);
>
> @@ -787,8 +808,18 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
> gmu_write(gmu, REG_A6XX_GMU_HFI_QTBL_ADDR, gmu->hfi.iova);
> gmu_write(gmu, REG_A6XX_GMU_HFI_QTBL_INFO, 1);
>
> + if (adreno_is_a7xx(adreno_gpu)) {
> + fence_range_upper = 0x32;
> + fence_range_lower = 0x8a0;
> + } else {
> + fence_range_upper = 0xa;
> + fence_range_lower = 0xa0;
> + }
> +
> gmu_write(gmu, REG_A6XX_GMU_AHB_FENCE_RANGE_0,
> - (1 << 31) | (0xa << 18) | (0xa0));
> + BIT(31) |
> + FIELD_PREP(GENMASK(30, 18), fence_range_upper) |
> + FIELD_PREP(GENMASK(17, 0), fence_range_lower));
This fails on arm32 because of the missing #include <linux/bitfields.h>
>
> /*
> * Snapshots toggle the NMI bit which will result in a jump to the NMI
--
With best wishes
Dmitry
^ permalink raw reply [flat|nested] 13+ messages in thread
* [PATCH v3 06/10] drm/msm/a6xx: Send ACD state to QMP at GMU resume
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
` (4 preceding siblings ...)
2023-08-23 12:55 ` [PATCH v3 05/10] drm/msm/a6xx: Add skeleton A7xx support Konrad Dybcio
@ 2023-08-23 12:55 ` Konrad Dybcio
2023-08-23 12:56 ` [PATCH v3 07/10] drm/msm/a6xx: Mostly implement A7xx gpu_state Konrad Dybcio
` (3 subsequent siblings)
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:55 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong
The QMP mailbox expects to be notified of the ACD (Adaptive Clock
Distribution) state. Get a handle to the mailbox at probe time and
poke it at GMU resume.
Since we don't fully support ACD yet, hardcode the message to "val: 0"
(state = disabled).
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 21 +++++++++++++++++++++
drivers/gpu/drm/msm/adreno/a6xx_gmu.h | 3 +++
2 files changed, 24 insertions(+)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
index 75984260898e..17e1e72f5d7d 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
@@ -980,11 +980,13 @@ static void a6xx_gmu_set_initial_bw(struct msm_gpu *gpu, struct a6xx_gmu *gmu)
dev_pm_opp_put(gpu_opp);
}
+#define GMU_ACD_STATE_MSG_LEN 36
int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
{
struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
struct msm_gpu *gpu = &adreno_gpu->base;
struct a6xx_gmu *gmu = &a6xx_gpu->gmu;
+ char buf[GMU_ACD_STATE_MSG_LEN];
int status, ret;
if (WARN(!gmu->initialized, "The GMU is not set up yet\n"))
@@ -992,6 +994,18 @@ int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
gmu->hung = false;
+ /* Notify AOSS about the ACD state (unimplemented for now => disable it) */
+ if (!IS_ERR(gmu->qmp)) {
+ ret = snprintf(buf, sizeof(buf),
+ "{class: gpu, res: acd, val: %d}",
+ 0 /* Hardcode ACD to be disabled for now */);
+ WARN_ON(ret >= GMU_ACD_STATE_MSG_LEN);
+
+ ret = qmp_send(gmu->qmp, buf, sizeof(buf));
+ if (ret)
+ dev_err(gmu->dev, "failed to send GPU ACD state\n");
+ }
+
/* Turn on the resources */
pm_runtime_get_sync(gmu->dev);
@@ -1744,6 +1758,10 @@ int a6xx_gmu_init(struct a6xx_gpu *a6xx_gpu, struct device_node *node)
goto detach_cxpd;
}
+ gmu->qmp = qmp_get(gmu->dev);
+ if (IS_ERR(gmu->qmp) && adreno_is_a7xx(adreno_gpu))
+ return PTR_ERR(gmu->qmp);
+
init_completion(&gmu->pd_gate);
complete_all(&gmu->pd_gate);
gmu->pd_nb.notifier_call = cxpd_notifier_cb;
@@ -1767,6 +1785,9 @@ int a6xx_gmu_init(struct a6xx_gpu *a6xx_gpu, struct device_node *node)
return 0;
+ if (!IS_ERR_OR_NULL(gmu->qmp))
+ qmp_put(gmu->qmp);
+
detach_cxpd:
dev_pm_domain_detach(gmu->cxpd, false);
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.h b/drivers/gpu/drm/msm/adreno/a6xx_gmu.h
index 236f81a43caa..592b296aab22 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.h
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.h
@@ -8,6 +8,7 @@
#include <linux/iopoll.h>
#include <linux/interrupt.h>
#include <linux/notifier.h>
+#include <linux/soc/qcom/qcom_aoss.h>
#include "msm_drv.h"
#include "a6xx_hfi.h"
@@ -96,6 +97,8 @@ struct a6xx_gmu {
/* For power domain callback */
struct notifier_block pd_nb;
struct completion pd_gate;
+
+ struct qmp *qmp;
};
static inline u32 gmu_read(struct a6xx_gmu *gmu, u32 offset)
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH v3 07/10] drm/msm/a6xx: Mostly implement A7xx gpu_state
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
` (5 preceding siblings ...)
2023-08-23 12:55 ` [PATCH v3 06/10] drm/msm/a6xx: Send ACD state to QMP at GMU resume Konrad Dybcio
@ 2023-08-23 12:56 ` Konrad Dybcio
2023-08-23 12:56 ` [PATCH v3 08/10] drm/msm/a6xx: Add A730 support Konrad Dybcio
` (2 subsequent siblings)
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:56 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong
Provide the necessary alternations to mostly support state dumping on
A7xx. Newer GPUs will probably require more changes here. Crashdumper
and debugbus remain untested.
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c | 52 +++++++++++++++++++++++-
drivers/gpu/drm/msm/adreno/a6xx_gpu_state.h | 61 ++++++++++++++++++++++++++++-
2 files changed, 110 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c
index 4e5d650578c6..18be2d3bde09 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c
@@ -948,6 +948,18 @@ static u32 a6xx_get_cp_roq_size(struct msm_gpu *gpu)
return gpu_read(gpu, REG_A6XX_CP_ROQ_THRESHOLDS_2) >> 14;
}
+static u32 a7xx_get_cp_roq_size(struct msm_gpu *gpu)
+{
+ /*
+ * The value at CP_ROQ_THRESHOLDS_2[20:31] is in 4dword units.
+ * That register however is not directly accessible from APSS on A7xx.
+ * Program the SQE_UCODE_DBG_ADDR with offset=0x70d3 and read the value.
+ */
+ gpu_write(gpu, REG_A6XX_CP_SQE_UCODE_DBG_ADDR, 0x70d3);
+
+ return 4 * (gpu_read(gpu, REG_A6XX_CP_SQE_UCODE_DBG_DATA) >> 20);
+}
+
/* Read a block of data from an indexed register pair */
static void a6xx_get_indexed_regs(struct msm_gpu *gpu,
struct a6xx_gpu_state *a6xx_state,
@@ -1019,8 +1031,40 @@ static void a6xx_get_indexed_registers(struct msm_gpu *gpu,
/* Restore the size in the hardware */
gpu_write(gpu, REG_A6XX_CP_MEM_POOL_SIZE, mempool_size);
+}
+
+static void a7xx_get_indexed_registers(struct msm_gpu *gpu,
+ struct a6xx_gpu_state *a6xx_state)
+{
+ int i, indexed_count, mempool_count;
+
+ indexed_count = ARRAY_SIZE(a7xx_indexed_reglist);
+ mempool_count = ARRAY_SIZE(a7xx_cp_bv_mempool_indexed);
- a6xx_state->nr_indexed_regs = count;
+ a6xx_state->indexed_regs = state_kcalloc(a6xx_state,
+ indexed_count + mempool_count,
+ sizeof(*a6xx_state->indexed_regs));
+ if (!a6xx_state->indexed_regs)
+ return;
+
+ a6xx_state->nr_indexed_regs = indexed_count + mempool_count;
+
+ /* First read the common regs */
+ for (i = 0; i < indexed_count; i++)
+ a6xx_get_indexed_regs(gpu, a6xx_state, &a7xx_indexed_reglist[i],
+ &a6xx_state->indexed_regs[i]);
+
+ gpu_rmw(gpu, REG_A6XX_CP_CHICKEN_DBG, 0, BIT(2));
+ gpu_rmw(gpu, REG_A7XX_CP_BV_CHICKEN_DBG, 0, BIT(2));
+
+ /* Get the contents of the CP_BV mempool */
+ for (i = 0; i < mempool_count; i++)
+ a6xx_get_indexed_regs(gpu, a6xx_state, a7xx_cp_bv_mempool_indexed,
+ &a6xx_state->indexed_regs[indexed_count - 1 + i]);
+
+ gpu_rmw(gpu, REG_A6XX_CP_CHICKEN_DBG, BIT(2), 0);
+ gpu_rmw(gpu, REG_A7XX_CP_BV_CHICKEN_DBG, BIT(2), 0);
+ return;
}
struct msm_gpu_state *a6xx_gpu_state_get(struct msm_gpu *gpu)
@@ -1056,6 +1100,12 @@ struct msm_gpu_state *a6xx_gpu_state_get(struct msm_gpu *gpu)
return &a6xx_state->base;
/* Get the banks of indexed registers */
+ if (adreno_is_a7xx(adreno_gpu)) {
+ a7xx_get_indexed_registers(gpu, a6xx_state);
+ /* Further codeflow is untested on A7xx. */
+ return &a6xx_state->base;
+ }
+
a6xx_get_indexed_registers(gpu, a6xx_state);
/*
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.h b/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.h
index e788ed72eb0d..8d7e6f26480a 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.h
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.h
@@ -338,6 +338,28 @@ static const struct a6xx_registers a6xx_vbif_reglist =
static const struct a6xx_registers a6xx_gbif_reglist =
REGS(a6xx_gbif_registers, 0, 0);
+static const u32 a7xx_ahb_registers[] = {
+ /* RBBM_STATUS */
+ 0x210, 0x210,
+ /* RBBM_STATUS2-3 */
+ 0x212, 0x213,
+};
+
+static const u32 a7xx_gbif_registers[] = {
+ 0x3c00, 0x3c0b,
+ 0x3c40, 0x3c42,
+ 0x3c45, 0x3c47,
+ 0x3c49, 0x3c4a,
+ 0x3cc0, 0x3cd1,
+};
+
+static const struct a6xx_registers a7xx_ahb_reglist[] = {
+ REGS(a7xx_ahb_registers, 0, 0),
+};
+
+static const struct a6xx_registers a7xx_gbif_reglist =
+ REGS(a7xx_gbif_registers, 0, 0);
+
static const u32 a6xx_gmu_gx_registers[] = {
/* GMU GX */
0x0000, 0x0000, 0x0010, 0x0013, 0x0016, 0x0016, 0x0018, 0x001b,
@@ -384,14 +406,17 @@ static const struct a6xx_registers a6xx_gmu_reglist[] = {
};
static u32 a6xx_get_cp_roq_size(struct msm_gpu *gpu);
+static u32 a7xx_get_cp_roq_size(struct msm_gpu *gpu);
-static struct a6xx_indexed_registers {
+struct a6xx_indexed_registers {
const char *name;
u32 addr;
u32 data;
u32 count;
u32 (*count_fn)(struct msm_gpu *gpu);
-} a6xx_indexed_reglist[] = {
+};
+
+static struct a6xx_indexed_registers a6xx_indexed_reglist[] = {
{ "CP_SQE_STAT", REG_A6XX_CP_SQE_STAT_ADDR,
REG_A6XX_CP_SQE_STAT_DATA, 0x33, NULL },
{ "CP_DRAW_STATE", REG_A6XX_CP_DRAW_STATE_ADDR,
@@ -402,11 +427,43 @@ static struct a6xx_indexed_registers {
REG_A6XX_CP_ROQ_DBG_DATA, 0, a6xx_get_cp_roq_size},
};
+static struct a6xx_indexed_registers a7xx_indexed_reglist[] = {
+ { "CP_SQE_STAT", REG_A6XX_CP_SQE_STAT_ADDR,
+ REG_A6XX_CP_SQE_STAT_DATA, 0x33, NULL },
+ { "CP_DRAW_STATE", REG_A6XX_CP_DRAW_STATE_ADDR,
+ REG_A6XX_CP_DRAW_STATE_DATA, 0x100, NULL },
+ { "CP_UCODE_DBG_DATA", REG_A6XX_CP_SQE_UCODE_DBG_ADDR,
+ REG_A6XX_CP_SQE_UCODE_DBG_DATA, 0x8000, NULL },
+ { "CP_BV_SQE_STAT_ADDR", REG_A7XX_CP_BV_SQE_STAT_ADDR,
+ REG_A7XX_CP_BV_SQE_STAT_DATA, 0x33, NULL },
+ { "CP_BV_DRAW_STATE_ADDR", REG_A7XX_CP_BV_DRAW_STATE_ADDR,
+ REG_A7XX_CP_BV_DRAW_STATE_DATA, 0x100, NULL },
+ { "CP_BV_SQE_UCODE_DBG_ADDR", REG_A7XX_CP_BV_SQE_UCODE_DBG_ADDR,
+ REG_A7XX_CP_BV_SQE_UCODE_DBG_DATA, 0x8000, NULL },
+ { "CP_SQE_AC_STAT_ADDR", REG_A7XX_CP_SQE_AC_STAT_ADDR,
+ REG_A7XX_CP_SQE_AC_STAT_DATA, 0x33, NULL },
+ { "CP_LPAC_DRAW_STATE_ADDR", REG_A7XX_CP_LPAC_DRAW_STATE_ADDR,
+ REG_A7XX_CP_LPAC_DRAW_STATE_DATA, 0x100, NULL },
+ { "CP_SQE_AC_UCODE_DBG_ADDR", REG_A7XX_CP_SQE_AC_UCODE_DBG_ADDR,
+ REG_A7XX_CP_SQE_AC_UCODE_DBG_DATA, 0x8000, NULL },
+ { "CP_LPAC_FIFO_DBG_ADDR", REG_A7XX_CP_LPAC_FIFO_DBG_ADDR,
+ REG_A7XX_CP_LPAC_FIFO_DBG_DATA, 0x40, NULL },
+ { "CP_ROQ", REG_A6XX_CP_ROQ_DBG_ADDR,
+ REG_A6XX_CP_ROQ_DBG_DATA, 0, a7xx_get_cp_roq_size },
+};
+
static struct a6xx_indexed_registers a6xx_cp_mempool_indexed = {
"CP_MEMPOOL", REG_A6XX_CP_MEM_POOL_DBG_ADDR,
REG_A6XX_CP_MEM_POOL_DBG_DATA, 0x2060, NULL,
};
+static struct a6xx_indexed_registers a7xx_cp_bv_mempool_indexed[] = {
+ { "CP_MEMPOOL", REG_A6XX_CP_MEM_POOL_DBG_ADDR,
+ REG_A6XX_CP_MEM_POOL_DBG_DATA, 0x2100, NULL },
+ { "CP_BV_MEMPOOL", REG_A7XX_CP_BV_MEM_POOL_DBG_ADDR,
+ REG_A7XX_CP_BV_MEM_POOL_DBG_DATA, 0x2100, NULL },
+};
+
#define DEBUGBUS(_id, _count) { .id = _id, .name = #_id, .count = _count }
static const struct a6xx_debugbus_block {
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH v3 08/10] drm/msm/a6xx: Add A730 support
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
` (6 preceding siblings ...)
2023-08-23 12:56 ` [PATCH v3 07/10] drm/msm/a6xx: Mostly implement A7xx gpu_state Konrad Dybcio
@ 2023-08-23 12:56 ` Konrad Dybcio
2023-08-23 12:56 ` [PATCH v3 09/10] drm/msm/a6xx: Add A740 support Konrad Dybcio
2023-08-23 12:56 ` [PATCH v3 10/10] drm/msm/a6xx: Poll for GBIF unhalt status in hw_init Konrad Dybcio
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:56 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong
Add support for Adreno 730, also known as GEN7_0_x, found on SM8450.
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 126 ++++++++++++++++++++++++++++-
drivers/gpu/drm/msm/adreno/a6xx_hfi.c | 61 ++++++++++++++
drivers/gpu/drm/msm/adreno/adreno_device.c | 13 +++
drivers/gpu/drm/msm/adreno/adreno_gpu.h | 7 +-
4 files changed, 203 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
index 61ce8d053355..522043883290 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
@@ -837,6 +837,63 @@ const struct adreno_reglist a690_hwcg[] = {
{}
};
+const struct adreno_reglist a730_hwcg[] = {
+ { REG_A6XX_RBBM_CLOCK_CNTL_SP0, 0x02222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_SP0, 0x02022222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_SP0, 0x0000f3cf },
+ { REG_A6XX_RBBM_CLOCK_DELAY_SP0, 0x00000080 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_TP0, 0x22222220 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_TP0, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL3_TP0, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL4_TP0, 0x00222222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_TP0, 0x77777777 },
+ { REG_A6XX_RBBM_CLOCK_HYST2_TP0, 0x77777777 },
+ { REG_A6XX_RBBM_CLOCK_HYST3_TP0, 0x77777777 },
+ { REG_A6XX_RBBM_CLOCK_HYST4_TP0, 0x00077777 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_TP0, 0x11111111 },
+ { REG_A6XX_RBBM_CLOCK_DELAY2_TP0, 0x11111111 },
+ { REG_A6XX_RBBM_CLOCK_DELAY3_TP0, 0x11111111 },
+ { REG_A6XX_RBBM_CLOCK_DELAY4_TP0, 0x00011111 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_UCHE, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_UCHE, 0x00000004 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_UCHE, 0x00000002 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_RB0, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_RB0, 0x01002222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_CCU0, 0x00002220 },
+ { REG_A6XX_RBBM_CLOCK_HYST_RB_CCU0, 0x44000f00 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_RAC, 0x25222022 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_RAC, 0x00555555 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_RAC, 0x00000011 },
+ { REG_A6XX_RBBM_CLOCK_HYST_RAC, 0x00440044 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_TSE_RAS_RBBM, 0x04222222 },
+ { REG_A7XX_RBBM_CLOCK_MODE2_GRAS, 0x00000222 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_GRAS, 0x00222222 },
+ { REG_A6XX_RBBM_CLOCK_MODE_GPC, 0x02222223 },
+ { REG_A6XX_RBBM_CLOCK_MODE_VFD, 0x00002222 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_GPC, 0x00222222 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_VFD, 0x00002222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_TSE_RAS_RBBM, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_HYST_GPC, 0x04104004 },
+ { REG_A6XX_RBBM_CLOCK_HYST_VFD, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_TSE_RAS_RBBM, 0x00004000 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_GPC, 0x00000200 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_VFD, 0x00002222 },
+ { REG_A6XX_RBBM_CLOCK_MODE_HLSQ, 0x00002222 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_HLSQ, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_HYST_HLSQ, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_HLSQ_2, 0x00000002 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_LRZ, 0x55555552 },
+ { REG_A7XX_RBBM_CLOCK_MODE_CP, 0x00000223 },
+ { REG_A6XX_RBBM_CLOCK_CNTL, 0x8aa8aa82 },
+ { REG_A6XX_RBBM_ISDB_CNT, 0x00000182 },
+ { REG_A6XX_RBBM_RAC_THRESHOLD_CNT, 0x00000000 },
+ { REG_A6XX_RBBM_SP_HYST_CNT, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_GMU_GX, 0x00000222 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_GMU_GX, 0x00000111 },
+ { REG_A6XX_RBBM_CLOCK_HYST_GMU_GX, 0x00000555 },
+ {},
+};
+
static void a6xx_set_hwcg(struct msm_gpu *gpu, bool state)
{
struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
@@ -1048,6 +1105,59 @@ static const u32 a690_protect[] = {
A6XX_PROTECT_NORDWR(0x11c00, 0x00000), /*note: infiite range */
};
+static const u32 a730_protect[] = {
+ A6XX_PROTECT_RDONLY(0x00000, 0x04ff),
+ A6XX_PROTECT_RDONLY(0x0050b, 0x0058),
+ A6XX_PROTECT_NORDWR(0x0050e, 0x0000),
+ A6XX_PROTECT_NORDWR(0x00510, 0x0000),
+ A6XX_PROTECT_NORDWR(0x00534, 0x0000),
+ A6XX_PROTECT_RDONLY(0x005fb, 0x009d),
+ A6XX_PROTECT_NORDWR(0x00699, 0x01e9),
+ A6XX_PROTECT_NORDWR(0x008a0, 0x0008),
+ A6XX_PROTECT_NORDWR(0x008ab, 0x0024),
+ /* 0x008d0-0x008dd are unprotected on purpose for tools like perfetto */
+ A6XX_PROTECT_RDONLY(0x008de, 0x0154),
+ A6XX_PROTECT_NORDWR(0x00900, 0x004d),
+ A6XX_PROTECT_NORDWR(0x0098d, 0x00b2),
+ A6XX_PROTECT_NORDWR(0x00a41, 0x01be),
+ A6XX_PROTECT_NORDWR(0x00df0, 0x0001),
+ A6XX_PROTECT_NORDWR(0x00e01, 0x0000),
+ A6XX_PROTECT_NORDWR(0x00e07, 0x0008),
+ A6XX_PROTECT_NORDWR(0x03c00, 0x00c3),
+ A6XX_PROTECT_RDONLY(0x03cc4, 0x1fff),
+ A6XX_PROTECT_NORDWR(0x08630, 0x01cf),
+ A6XX_PROTECT_NORDWR(0x08e00, 0x0000),
+ A6XX_PROTECT_NORDWR(0x08e08, 0x0000),
+ A6XX_PROTECT_NORDWR(0x08e50, 0x001f),
+ A6XX_PROTECT_NORDWR(0x08e80, 0x0280),
+ A6XX_PROTECT_NORDWR(0x09624, 0x01db),
+ A6XX_PROTECT_NORDWR(0x09e40, 0x0000),
+ A6XX_PROTECT_NORDWR(0x09e64, 0x000d),
+ A6XX_PROTECT_NORDWR(0x09e78, 0x0187),
+ A6XX_PROTECT_NORDWR(0x0a630, 0x01cf),
+ A6XX_PROTECT_NORDWR(0x0ae02, 0x0000),
+ A6XX_PROTECT_NORDWR(0x0ae50, 0x000f),
+ A6XX_PROTECT_NORDWR(0x0ae66, 0x0003),
+ A6XX_PROTECT_NORDWR(0x0ae6f, 0x0003),
+ A6XX_PROTECT_NORDWR(0x0b604, 0x0003),
+ A6XX_PROTECT_NORDWR(0x0ec00, 0x0fff),
+ A6XX_PROTECT_RDONLY(0x0fc00, 0x1fff),
+ A6XX_PROTECT_NORDWR(0x18400, 0x0053),
+ A6XX_PROTECT_RDONLY(0x18454, 0x0004),
+ A6XX_PROTECT_NORDWR(0x18459, 0x1fff),
+ A6XX_PROTECT_NORDWR(0x1a459, 0x1fff),
+ A6XX_PROTECT_NORDWR(0x1c459, 0x1fff),
+ A6XX_PROTECT_NORDWR(0x1f400, 0x0443),
+ A6XX_PROTECT_RDONLY(0x1f844, 0x007b),
+ A6XX_PROTECT_NORDWR(0x1f860, 0x0000),
+ A6XX_PROTECT_NORDWR(0x1f878, 0x002a),
+ /* CP_PROTECT_REG[44, 46] are left untouched! */
+ 0,
+ 0,
+ 0,
+ A6XX_PROTECT_NORDWR(0x1f8c0, 0x00000),
+};
+
static void a6xx_set_cp_protect(struct msm_gpu *gpu)
{
struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
@@ -1069,6 +1179,11 @@ static void a6xx_set_cp_protect(struct msm_gpu *gpu)
count = ARRAY_SIZE(a660_protect);
count_max = 48;
BUILD_BUG_ON(ARRAY_SIZE(a660_protect) > 48);
+ } else if (adreno_is_a730(adreno_gpu)) {
+ regs = a730_protect;
+ count = ARRAY_SIZE(a730_protect);
+ count_max = 48;
+ BUILD_BUG_ON(ARRAY_SIZE(a730_protect) > 48);
} else {
regs = a6xx_protect;
count = ARRAY_SIZE(a6xx_protect);
@@ -1135,7 +1250,9 @@ static void a6xx_set_ubwc_config(struct msm_gpu *gpu)
if (adreno_is_a640_family(adreno_gpu))
amsbc = 1;
- if (adreno_is_a650(adreno_gpu) || adreno_is_a660(adreno_gpu)) {
+ if (adreno_is_a650(adreno_gpu) ||
+ adreno_is_a660(adreno_gpu) ||
+ adreno_is_a730(adreno_gpu)) {
/* TODO: get ddr type from bootloader and use 2 for LPDDR4 */
hbb_lo = 3;
amsbc = 1;
@@ -1516,7 +1633,8 @@ static int hw_init(struct msm_gpu *gpu)
gpu_write64(gpu, REG_A6XX_UCHE_WRITE_THRU_BASE, 0x0001fffffffff000llu);
}
- if (!adreno_is_a650_family(adreno_gpu)) {
+ if (!(adreno_is_a650_family(adreno_gpu) ||
+ adreno_is_a730(adreno_gpu))) {
/* Set the GMEM VA range [0x100000:0x100000 + gpu->gmem - 1] */
gpu_write64(gpu, REG_A6XX_UCHE_GMEM_RANGE_MIN, 0x00100000);
@@ -1586,7 +1704,9 @@ static int hw_init(struct msm_gpu *gpu)
a6xx_set_ubwc_config(gpu);
/* Enable fault detection */
- if (adreno_is_a619(adreno_gpu))
+ if (adreno_is_a730(adreno_gpu))
+ gpu_write(gpu, REG_A6XX_RBBM_INTERFACE_HANG_INT_CNTL, (1 << 30) | 0xcfffff);
+ else if (adreno_is_a619(adreno_gpu))
gpu_write(gpu, REG_A6XX_RBBM_INTERFACE_HANG_INT_CNTL, (1 << 30) | 0x3fffff);
else if (adreno_is_a610(adreno_gpu))
gpu_write(gpu, REG_A6XX_RBBM_INTERFACE_HANG_INT_CNTL, (1 << 30) | 0x3ffff);
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_hfi.c b/drivers/gpu/drm/msm/adreno/a6xx_hfi.c
index 25b235b49ebc..3865cd44523c 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_hfi.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_hfi.c
@@ -5,6 +5,8 @@
#include <linux/circ_buf.h>
#include <linux/list.h>
+#include <soc/qcom/cmd-db.h>
+
#include "a6xx_gmu.h"
#include "a6xx_gmu.xml.h"
#include "a6xx_gpu.h"
@@ -506,6 +508,63 @@ static void adreno_7c3_build_bw_table(struct a6xx_hfi_msg_bw_table *msg)
msg->cnoc_cmds_data[0][0] = 0x40000000;
msg->cnoc_cmds_data[1][0] = 0x60000001;
}
+
+static void a730_build_bw_table(struct a6xx_hfi_msg_bw_table *msg)
+{
+ msg->bw_level_num = 12;
+
+ msg->ddr_cmds_num = 3;
+ msg->ddr_wait_bitmask = 0x7;
+
+ msg->ddr_cmds_addrs[0] = cmd_db_read_addr("SH0");
+ msg->ddr_cmds_addrs[1] = cmd_db_read_addr("MC0");
+ msg->ddr_cmds_addrs[2] = cmd_db_read_addr("ACV");
+
+ msg->ddr_cmds_data[0][0] = 0x40000000;
+ msg->ddr_cmds_data[0][1] = 0x40000000;
+ msg->ddr_cmds_data[0][2] = 0x40000000;
+ msg->ddr_cmds_data[1][0] = 0x600002e8;
+ msg->ddr_cmds_data[1][1] = 0x600003d0;
+ msg->ddr_cmds_data[1][2] = 0x60000008;
+ msg->ddr_cmds_data[2][0] = 0x6000068d;
+ msg->ddr_cmds_data[2][1] = 0x6000089a;
+ msg->ddr_cmds_data[2][2] = 0x60000008;
+ msg->ddr_cmds_data[3][0] = 0x600007f2;
+ msg->ddr_cmds_data[3][1] = 0x60000a6e;
+ msg->ddr_cmds_data[3][2] = 0x60000008;
+ msg->ddr_cmds_data[4][0] = 0x600009e5;
+ msg->ddr_cmds_data[4][1] = 0x60000cfd;
+ msg->ddr_cmds_data[4][2] = 0x60000008;
+ msg->ddr_cmds_data[5][0] = 0x60000b29;
+ msg->ddr_cmds_data[5][1] = 0x60000ea6;
+ msg->ddr_cmds_data[5][2] = 0x60000008;
+ msg->ddr_cmds_data[6][0] = 0x60001698;
+ msg->ddr_cmds_data[6][1] = 0x60001da8;
+ msg->ddr_cmds_data[6][2] = 0x60000008;
+ msg->ddr_cmds_data[7][0] = 0x600018d2;
+ msg->ddr_cmds_data[7][1] = 0x60002093;
+ msg->ddr_cmds_data[7][2] = 0x60000008;
+ msg->ddr_cmds_data[8][0] = 0x60001e66;
+ msg->ddr_cmds_data[8][1] = 0x600027e6;
+ msg->ddr_cmds_data[8][2] = 0x60000008;
+ msg->ddr_cmds_data[9][0] = 0x600027c2;
+ msg->ddr_cmds_data[9][1] = 0x6000342f;
+ msg->ddr_cmds_data[9][2] = 0x60000008;
+ msg->ddr_cmds_data[10][0] = 0x60002e71;
+ msg->ddr_cmds_data[10][1] = 0x60003cf5;
+ msg->ddr_cmds_data[10][2] = 0x60000008;
+ msg->ddr_cmds_data[11][0] = 0x600030ae;
+ msg->ddr_cmds_data[11][1] = 0x60003fe5;
+ msg->ddr_cmds_data[11][2] = 0x60000008;
+
+ msg->cnoc_cmds_num = 1;
+ msg->cnoc_wait_bitmask = 0x1;
+
+ msg->cnoc_cmds_addrs[0] = cmd_db_read_addr("CN0");
+ msg->cnoc_cmds_data[0][0] = 0x40000000;
+ msg->cnoc_cmds_data[1][0] = 0x60000001;
+}
+
static void a6xx_build_bw_table(struct a6xx_hfi_msg_bw_table *msg)
{
/* Send a single "off" entry since the 630 GMU doesn't do bus scaling */
@@ -564,6 +623,8 @@ static int a6xx_hfi_send_bw_table(struct a6xx_gmu *gmu)
a660_build_bw_table(&msg);
else if (adreno_is_a690(adreno_gpu))
a690_build_bw_table(&msg);
+ else if (adreno_is_a730(adreno_gpu))
+ a730_build_bw_table(&msg);
else
a6xx_build_bw_table(&msg);
diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c
index 575e7c56219f..44599f97db20 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_device.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_device.c
@@ -490,6 +490,19 @@ static const struct adreno_info gpulist[] = {
.zapfw = "a690_zap.mdt",
.hwcg = a690_hwcg,
.address_space_size = SZ_16G,
+ }, {
+ .chip_ids = ADRENO_CHIP_IDS(0x07030001),
+ .family = ADRENO_7XX_GEN1,
+ .fw = {
+ [ADRENO_FW_SQE] = "a730_sqe.fw",
+ [ADRENO_FW_GMU] = "gmu_gen70000.bin",
+ },
+ .gmem = SZ_2M,
+ .inactive_period = DRM_MSM_INACTIVE_PERIOD,
+ .init = a6xx_gpu_init,
+ .zapfw = "a730_zap.mdt",
+ .hwcg = a730_hwcg,
+ .address_space_size = SZ_16G,
},
};
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index ad07e788a69c..fe38a1fc962b 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -76,7 +76,7 @@ struct adreno_reglist {
};
extern const struct adreno_reglist a612_hwcg[], a615_hwcg[], a630_hwcg[], a640_hwcg[], a650_hwcg[];
-extern const struct adreno_reglist a660_hwcg[], a690_hwcg[];
+extern const struct adreno_reglist a660_hwcg[], a690_hwcg[], a730_hwcg[];
struct adreno_speedbin {
uint16_t fuse;
@@ -403,6 +403,11 @@ static inline int adreno_is_a640_family(const struct adreno_gpu *gpu)
return gpu->info->family == ADRENO_6XX_GEN2;
}
+static inline int adreno_is_a730(struct adreno_gpu *gpu)
+{
+ return gpu->info->chip_ids[0] == 0x07030001;
+}
+
static inline int adreno_is_a7xx(struct adreno_gpu *gpu)
{
/* Update with non-fake (i.e. non-A702) Gen 7 GPUs */
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH v3 09/10] drm/msm/a6xx: Add A740 support
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
` (7 preceding siblings ...)
2023-08-23 12:56 ` [PATCH v3 08/10] drm/msm/a6xx: Add A730 support Konrad Dybcio
@ 2023-08-23 12:56 ` Konrad Dybcio
2023-09-12 12:52 ` Konrad Dybcio
2023-08-23 12:56 ` [PATCH v3 10/10] drm/msm/a6xx: Poll for GBIF unhalt status in hw_init Konrad Dybcio
9 siblings, 1 reply; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:56 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong
A740 builds upon the A730 IP, shuffling some values and registers
around. More differences will appear when things like BCL are
implemented.
adreno_is_a740_family is added in preparation for more A7xx GPUs,
the logic checks will be valid resulting in smaller diffs.
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 88 +++++++++++++++++++++---------
drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 82 +++++++++++++++++++++++++---
drivers/gpu/drm/msm/adreno/a6xx_hfi.c | 27 +++++++++
drivers/gpu/drm/msm/adreno/adreno_device.c | 17 ++++++
drivers/gpu/drm/msm/adreno/adreno_gpu.c | 6 +-
drivers/gpu/drm/msm/adreno/adreno_gpu.h | 19 ++++++-
6 files changed, 201 insertions(+), 38 deletions(-)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
index 17e1e72f5d7d..14ba407e7fe0 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
@@ -516,6 +516,7 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
struct platform_device *pdev = to_platform_device(gmu->dev);
void __iomem *pdcptr = a6xx_gmu_get_mmio(pdev, "gmu_pdc");
+ u32 seqmem0_drv0_reg = REG_A6XX_RSCC_SEQ_MEM_0_DRV0;
void __iomem *seqptr = NULL;
uint32_t pdc_address_offset;
bool pdc_in_aop = false;
@@ -549,21 +550,26 @@ static void a6xx_gmu_rpmh_init(struct a6xx_gmu *gmu)
gmu_write_rscc(gmu, REG_A6XX_RSCC_HIDDEN_TCS_CMD0_ADDR, 0);
gmu_write_rscc(gmu, REG_A6XX_RSCC_HIDDEN_TCS_CMD0_DATA + 2, 0);
gmu_write_rscc(gmu, REG_A6XX_RSCC_HIDDEN_TCS_CMD0_ADDR + 2, 0);
- gmu_write_rscc(gmu, REG_A6XX_RSCC_HIDDEN_TCS_CMD0_DATA + 4, 0x80000000);
+ gmu_write_rscc(gmu, REG_A6XX_RSCC_HIDDEN_TCS_CMD0_DATA + 4,
+ adreno_is_a740_family(adreno_gpu) ? 0x80000021 : 0x80000000);
gmu_write_rscc(gmu, REG_A6XX_RSCC_HIDDEN_TCS_CMD0_ADDR + 4, 0);
gmu_write_rscc(gmu, REG_A6XX_RSCC_OVERRIDE_START_ADDR, 0);
gmu_write_rscc(gmu, REG_A6XX_RSCC_PDC_SEQ_START_ADDR, 0x4520);
gmu_write_rscc(gmu, REG_A6XX_RSCC_PDC_MATCH_VALUE_LO, 0x4510);
gmu_write_rscc(gmu, REG_A6XX_RSCC_PDC_MATCH_VALUE_HI, 0x4514);
+ /* The second spin of A7xx GPUs messed with some register offsets.. */
+ if (adreno_is_a740_family(adreno_gpu))
+ seqmem0_drv0_reg = REG_A7XX_RSCC_SEQ_MEM_0_DRV0_A740;
+
/* Load RSC sequencer uCode for sleep and wakeup */
if (adreno_is_a650_family(adreno_gpu) ||
adreno_is_a7xx(adreno_gpu)) {
- gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0, 0xeaaae5a0);
- gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 1, 0xe1a1ebab);
- gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 2, 0xa2e0a581);
- gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 3, 0xecac82e2);
- gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 4, 0x0020edad);
+ gmu_write_rscc(gmu, seqmem0_drv0_reg, 0xeaaae5a0);
+ gmu_write_rscc(gmu, seqmem0_drv0_reg + 1, 0xe1a1ebab);
+ gmu_write_rscc(gmu, seqmem0_drv0_reg + 2, 0xa2e0a581);
+ gmu_write_rscc(gmu, seqmem0_drv0_reg + 3, 0xecac82e2);
+ gmu_write_rscc(gmu, seqmem0_drv0_reg + 4, 0x0020edad);
} else {
gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0, 0xa7a506a0);
gmu_write_rscc(gmu, REG_A6XX_RSCC_SEQ_MEM_0_DRV0 + 1, 0xa1e6a6e7);
@@ -767,8 +773,8 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
u32 fence_range_lower, fence_range_upper;
+ u32 chipid, chipid_min = 0;
int ret;
- u32 chipid;
/* Vote veto for FAL10 */
if (adreno_is_a650_family(adreno_gpu) || adreno_is_a7xx(adreno_gpu)) {
@@ -827,16 +833,37 @@ static int a6xx_gmu_fw_start(struct a6xx_gmu *gmu, unsigned int state)
*/
gmu_write(gmu, REG_A6XX_GMU_CM3_CFG, 0x4052);
- /*
- * Note that the GMU has a slightly different layout for
- * chip_id, for whatever reason, so a bit of massaging
- * is needed. The upper 16b are the same, but minor and
- * patchid are packed in four bits each with the lower
- * 8b unused:
- */
- chipid = adreno_gpu->chip_id & 0xffff0000;
- chipid |= (adreno_gpu->chip_id << 4) & 0xf000; /* minor */
- chipid |= (adreno_gpu->chip_id << 8) & 0x0f00; /* patchid */
+ /* NOTE: A730 may also fall in this if-condition with a future GMU fw update. */
+ if (adreno_is_a7xx(adreno_gpu) && !adreno_is_a730(adreno_gpu)) {
+ /* A7xx GPUs have obfuscated chip IDs. Use constant maj = 7 */
+ chipid = FIELD_PREP(GENMASK(31, 24), 0x7);
+
+ /*
+ * The min part has a 1-1 mapping for each GPU SKU.
+ * This chipid that the GMU expects corresponds to the "GENX_Y_Z" naming,
+ * where X = major, Y = minor, Z = patchlevel, e.g. GEN7_2_1 for prod A740.
+ */
+ if (adreno_is_a740(adreno_gpu))
+ chipid_min = 2;
+ else
+ return -EINVAL;
+
+ chipid |= FIELD_PREP(GENMASK(23, 16), chipid_min);
+
+ /* Get the patchid (which may vary) from the device tree */
+ chipid |= FIELD_PREP(GENMASK(15, 8), adreno_patchid(adreno_gpu));
+ } else {
+ /*
+ * Note that the GMU has a slightly different layout for
+ * chip_id, for whatever reason, so a bit of massaging
+ * is needed. The upper 16b are the same, but minor and
+ * patchid are packed in four bits each with the lower
+ * 8b unused:
+ */
+ chipid = adreno_gpu->chip_id & 0xffff0000;
+ chipid |= (adreno_gpu->chip_id << 4) & 0xf000; /* minor */
+ chipid |= (adreno_gpu->chip_id << 8) & 0x0f00; /* patchid */
+ }
if (adreno_is_a7xx(adreno_gpu)) {
gmu_write(gmu, REG_A6XX_GMU_GENERAL_10, chipid);
@@ -899,17 +926,23 @@ static void a6xx_gmu_irq_disable(struct a6xx_gmu *gmu)
static void a6xx_gmu_rpmh_off(struct a6xx_gmu *gmu)
{
- u32 val;
+ struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
+ struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
+ u32 val, seqmem_off = 0;
+
+ /* The second spin of A7xx GPUs messed with some register offsets.. */
+ if (adreno_is_a740_family(adreno_gpu))
+ seqmem_off = 4;
/* Make sure there are no outstanding RPMh votes */
- gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS0_DRV0_STATUS, val,
- (val & 1), 100, 10000);
- gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS1_DRV0_STATUS, val,
- (val & 1), 100, 10000);
- gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS2_DRV0_STATUS, val,
- (val & 1), 100, 10000);
- gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS3_DRV0_STATUS, val,
- (val & 1), 100, 1000);
+ gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS0_DRV0_STATUS + seqmem_off,
+ val, (val & 1), 100, 10000);
+ gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS1_DRV0_STATUS + seqmem_off,
+ val, (val & 1), 100, 10000);
+ gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS2_DRV0_STATUS + seqmem_off,
+ val, (val & 1), 100, 10000);
+ gmu_poll_timeout_rscc(gmu, REG_A6XX_RSCC_TCS3_DRV0_STATUS + seqmem_off,
+ val, (val & 1), 100, 1000);
}
/* Force the GMU off in case it isn't responsive */
@@ -1019,7 +1052,8 @@ int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
/* Use a known rate to bring up the GMU */
clk_set_rate(gmu->core_clk, 200000000);
- clk_set_rate(gmu->hub_clk, 150000000);
+ clk_set_rate(gmu->hub_clk, adreno_is_a740_family(adreno_gpu) ?
+ 200000000 : 150000000);
ret = clk_bulk_prepare_enable(gmu->nr_clocks, gmu->clocks);
if (ret) {
pm_runtime_put(gmu->gxpd);
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
index 522043883290..2313620084b6 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
@@ -894,6 +894,64 @@ const struct adreno_reglist a730_hwcg[] = {
{},
};
+const struct adreno_reglist a740_hwcg[] = {
+ { REG_A6XX_RBBM_CLOCK_CNTL_SP0, 0x02222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_SP0, 0x22022222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_SP0, 0x003cf3cf },
+ { REG_A6XX_RBBM_CLOCK_DELAY_SP0, 0x00000080 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_TP0, 0x22222220 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_TP0, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL3_TP0, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL4_TP0, 0x00222222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_TP0, 0x77777777 },
+ { REG_A6XX_RBBM_CLOCK_HYST2_TP0, 0x77777777 },
+ { REG_A6XX_RBBM_CLOCK_HYST3_TP0, 0x77777777 },
+ { REG_A6XX_RBBM_CLOCK_HYST4_TP0, 0x00077777 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_TP0, 0x11111111 },
+ { REG_A6XX_RBBM_CLOCK_DELAY2_TP0, 0x11111111 },
+ { REG_A6XX_RBBM_CLOCK_DELAY3_TP0, 0x11111111 },
+ { REG_A6XX_RBBM_CLOCK_DELAY4_TP0, 0x00011111 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_UCHE, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_UCHE, 0x00222222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_UCHE, 0x00000444 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_UCHE, 0x00000222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_RB0, 0x22222222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_RB0, 0x01002222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_CCU0, 0x00002220 },
+ { REG_A6XX_RBBM_CLOCK_HYST_RB_CCU0, 0x44000f00 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_RAC, 0x25222022 },
+ { REG_A6XX_RBBM_CLOCK_CNTL2_RAC, 0x00555555 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_RAC, 0x00000011 },
+ { REG_A6XX_RBBM_CLOCK_HYST_RAC, 0x00440044 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_TSE_RAS_RBBM, 0x04222222 },
+ { REG_A7XX_RBBM_CLOCK_MODE2_GRAS, 0x00000222 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_GRAS, 0x00222222 },
+ { REG_A6XX_RBBM_CLOCK_MODE_GPC, 0x02222223 },
+ { REG_A6XX_RBBM_CLOCK_MODE_VFD, 0x00222222 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_GPC, 0x00222222 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_VFD, 0x00002222 },
+ { REG_A6XX_RBBM_CLOCK_HYST_TSE_RAS_RBBM, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_HYST_GPC, 0x04104004 },
+ { REG_A6XX_RBBM_CLOCK_HYST_VFD, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_TSE_RAS_RBBM, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_GPC, 0x00000200 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_VFD, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_MODE_HLSQ, 0x00002222 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_HLSQ, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_HYST_HLSQ, 0x00000000 },
+ { REG_A7XX_RBBM_CLOCK_MODE_BV_LRZ, 0x55555552 },
+ { REG_A7XX_RBBM_CLOCK_HYST2_VFD, 0x00000000 },
+ { REG_A7XX_RBBM_CLOCK_MODE_CP, 0x00000222 },
+ { REG_A6XX_RBBM_CLOCK_CNTL, 0x8aa8aa82 },
+ { REG_A6XX_RBBM_ISDB_CNT, 0x00000182 },
+ { REG_A6XX_RBBM_RAC_THRESHOLD_CNT, 0x00000000 },
+ { REG_A6XX_RBBM_SP_HYST_CNT, 0x00000000 },
+ { REG_A6XX_RBBM_CLOCK_CNTL_GMU_GX, 0x00000222 },
+ { REG_A6XX_RBBM_CLOCK_DELAY_GMU_GX, 0x00000111 },
+ { REG_A6XX_RBBM_CLOCK_HYST_GMU_GX, 0x00000555 },
+ {},
+};
+
static void a6xx_set_hwcg(struct msm_gpu *gpu, bool state)
{
struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
@@ -901,7 +959,7 @@ static void a6xx_set_hwcg(struct msm_gpu *gpu, bool state)
struct a6xx_gmu *gmu = &a6xx_gpu->gmu;
const struct adreno_reglist *reg;
unsigned int i;
- u32 val, clock_cntl_on;
+ u32 val, clock_cntl_on, cgc_mode;
if (!adreno_gpu->info->hwcg)
return;
@@ -914,8 +972,10 @@ static void a6xx_set_hwcg(struct msm_gpu *gpu, bool state)
clock_cntl_on = 0x8aa8aa82;
if (adreno_is_a7xx(adreno_gpu)) {
+ cgc_mode = adreno_is_a740_family(adreno_gpu) ? 0x20222 : 0x20000;
+
gmu_write(&a6xx_gpu->gmu, REG_A6XX_GPU_GMU_AO_GMU_CGC_MODE_CNTL,
- state ? 0x20000 : 0);
+ state ? cgc_mode : 0);
gmu_write(&a6xx_gpu->gmu, REG_A6XX_GPU_GMU_AO_GMU_CGC_DELAY_CNTL,
state ? 0x10111 : 0);
gmu_write(&a6xx_gpu->gmu, REG_A6XX_GPU_GMU_AO_GMU_CGC_HYST_CNTL,
@@ -1179,7 +1239,7 @@ static void a6xx_set_cp_protect(struct msm_gpu *gpu)
count = ARRAY_SIZE(a660_protect);
count_max = 48;
BUILD_BUG_ON(ARRAY_SIZE(a660_protect) > 48);
- } else if (adreno_is_a730(adreno_gpu)) {
+ } else if (adreno_is_a730(adreno_gpu) || adreno_is_a740(adreno_gpu)) {
regs = a730_protect;
count = ARRAY_SIZE(a730_protect);
count_max = 48;
@@ -1252,7 +1312,8 @@ static void a6xx_set_ubwc_config(struct msm_gpu *gpu)
if (adreno_is_a650(adreno_gpu) ||
adreno_is_a660(adreno_gpu) ||
- adreno_is_a730(adreno_gpu)) {
+ adreno_is_a730(adreno_gpu) ||
+ adreno_is_a740_family(adreno_gpu)) {
/* TODO: get ddr type from bootloader and use 2 for LPDDR4 */
hbb_lo = 3;
amsbc = 1;
@@ -1545,6 +1606,7 @@ static int hw_init(struct msm_gpu *gpu)
struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
struct a6xx_gpu *a6xx_gpu = to_a6xx_gpu(adreno_gpu);
struct a6xx_gmu *gmu = &a6xx_gpu->gmu;
+ u64 gmem_range_min;
int ret;
if (!adreno_has_gmu_wrapper(adreno_gpu)) {
@@ -1635,11 +1697,13 @@ static int hw_init(struct msm_gpu *gpu)
if (!(adreno_is_a650_family(adreno_gpu) ||
adreno_is_a730(adreno_gpu))) {
+ gmem_range_min = adreno_is_a740_family(adreno_gpu) ? SZ_16M : SZ_1M;
+
/* Set the GMEM VA range [0x100000:0x100000 + gpu->gmem - 1] */
- gpu_write64(gpu, REG_A6XX_UCHE_GMEM_RANGE_MIN, 0x00100000);
+ gpu_write64(gpu, REG_A6XX_UCHE_GMEM_RANGE_MIN, gmem_range_min);
gpu_write64(gpu, REG_A6XX_UCHE_GMEM_RANGE_MAX,
- 0x00100000 + adreno_gpu->info->gmem - 1);
+ gmem_range_min + adreno_gpu->info->gmem - 1);
}
if (adreno_is_a7xx(adreno_gpu))
@@ -1704,7 +1768,8 @@ static int hw_init(struct msm_gpu *gpu)
a6xx_set_ubwc_config(gpu);
/* Enable fault detection */
- if (adreno_is_a730(adreno_gpu))
+ if (adreno_is_a730(adreno_gpu) ||
+ adreno_is_a740_family(adreno_gpu))
gpu_write(gpu, REG_A6XX_RBBM_INTERFACE_HANG_INT_CNTL, (1 << 30) | 0xcfffff);
else if (adreno_is_a619(adreno_gpu))
gpu_write(gpu, REG_A6XX_RBBM_INTERFACE_HANG_INT_CNTL, (1 << 30) | 0x3fffff);
@@ -2796,7 +2861,8 @@ struct msm_gpu *a6xx_gpu_init(struct drm_device *dev)
!!(config->info->quirks & ADRENO_QUIRK_HAS_HW_APRIV);
/* gpu->info only gets assigned in adreno_gpu_init() */
- is_a7xx = config->info->family == ADRENO_7XX_GEN1;
+ is_a7xx = config->info->family == ADRENO_7XX_GEN1 ||
+ config->info->family == ADRENO_7XX_GEN2;
a6xx_llc_slices_init(pdev, a6xx_gpu, is_a7xx);
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_hfi.c b/drivers/gpu/drm/msm/adreno/a6xx_hfi.c
index 3865cd44523c..cdb3f6e74d3e 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_hfi.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_hfi.c
@@ -565,6 +565,31 @@ static void a730_build_bw_table(struct a6xx_hfi_msg_bw_table *msg)
msg->cnoc_cmds_data[1][0] = 0x60000001;
}
+static void a740_build_bw_table(struct a6xx_hfi_msg_bw_table *msg)
+{
+ msg->bw_level_num = 1;
+
+ msg->ddr_cmds_num = 3;
+ msg->ddr_wait_bitmask = 0x7;
+
+ msg->ddr_cmds_addrs[0] = cmd_db_read_addr("SH0");
+ msg->ddr_cmds_addrs[1] = cmd_db_read_addr("MC0");
+ msg->ddr_cmds_addrs[2] = cmd_db_read_addr("ACV");
+
+ msg->ddr_cmds_data[0][0] = 0x40000000;
+ msg->ddr_cmds_data[0][1] = 0x40000000;
+ msg->ddr_cmds_data[0][2] = 0x40000000;
+
+ /* TODO: add a proper dvfs table */
+
+ msg->cnoc_cmds_num = 1;
+ msg->cnoc_wait_bitmask = 0x1;
+
+ msg->cnoc_cmds_addrs[0] = cmd_db_read_addr("CN0");
+ msg->cnoc_cmds_data[0][0] = 0x40000000;
+ msg->cnoc_cmds_data[1][0] = 0x60000001;
+}
+
static void a6xx_build_bw_table(struct a6xx_hfi_msg_bw_table *msg)
{
/* Send a single "off" entry since the 630 GMU doesn't do bus scaling */
@@ -625,6 +650,8 @@ static int a6xx_hfi_send_bw_table(struct a6xx_gmu *gmu)
a690_build_bw_table(&msg);
else if (adreno_is_a730(adreno_gpu))
a730_build_bw_table(&msg);
+ else if (adreno_is_a740_family(adreno_gpu))
+ a740_build_bw_table(&msg);
else
a6xx_build_bw_table(&msg);
diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c
index 44599f97db20..a22d61190291 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_device.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_device.c
@@ -499,10 +499,27 @@ static const struct adreno_info gpulist[] = {
},
.gmem = SZ_2M,
.inactive_period = DRM_MSM_INACTIVE_PERIOD,
+ .quirks = ADRENO_QUIRK_HAS_CACHED_COHERENT |
+ ADRENO_QUIRK_HAS_HW_APRIV,
.init = a6xx_gpu_init,
.zapfw = "a730_zap.mdt",
.hwcg = a730_hwcg,
.address_space_size = SZ_16G,
+ }, {
+ .chip_ids = ADRENO_CHIP_IDS(0x43050a01), /* "C510v2" */
+ .family = ADRENO_7XX_GEN2,
+ .fw = {
+ [ADRENO_FW_SQE] = "a740_sqe.fw",
+ [ADRENO_FW_GMU] = "gmu_gen70200.bin",
+ },
+ .gmem = 3 * SZ_1M,
+ .inactive_period = DRM_MSM_INACTIVE_PERIOD,
+ .quirks = ADRENO_QUIRK_HAS_CACHED_COHERENT |
+ ADRENO_QUIRK_HAS_HW_APRIV,
+ .init = a6xx_gpu_init,
+ .zapfw = "a740_zap.mdt",
+ .hwcg = a740_hwcg,
+ .address_space_size = SZ_16G,
},
};
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index ea59724f8e41..3fe9fd240cc7 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -323,7 +323,11 @@ int adreno_get_param(struct msm_gpu *gpu, struct msm_file_private *ctx,
*value = adreno_gpu->info->gmem;
return 0;
case MSM_PARAM_GMEM_BASE:
- *value = !adreno_is_a650_family(adreno_gpu) ? 0x100000 : 0;
+ if (adreno_is_a650_family(adreno_gpu) ||
+ adreno_is_a740_family(adreno_gpu))
+ *value = 0;
+ else
+ *value = 0x100000;
return 0;
case MSM_PARAM_CHIP_ID:
*value = adreno_gpu->chip_id;
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index fe38a1fc962b..076d66dcaed5 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -47,6 +47,7 @@ enum adreno_family {
ADRENO_6XX_GEN3, /* a650 family */
ADRENO_6XX_GEN4, /* a660 family */
ADRENO_7XX_GEN1, /* a730 family */
+ ADRENO_7XX_GEN2, /* a740 family */
};
#define ADRENO_QUIRK_TWO_PASS_USE_WFI BIT(0)
@@ -76,7 +77,7 @@ struct adreno_reglist {
};
extern const struct adreno_reglist a612_hwcg[], a615_hwcg[], a630_hwcg[], a640_hwcg[], a650_hwcg[];
-extern const struct adreno_reglist a660_hwcg[], a690_hwcg[], a730_hwcg[];
+extern const struct adreno_reglist a660_hwcg[], a690_hwcg[], a730_hwcg[], a740_hwcg[];
struct adreno_speedbin {
uint16_t fuse;
@@ -408,10 +409,24 @@ static inline int adreno_is_a730(struct adreno_gpu *gpu)
return gpu->info->chip_ids[0] == 0x07030001;
}
+static inline int adreno_is_a740(struct adreno_gpu *gpu)
+{
+ return gpu->info->chip_ids[0] == 0x43050a01;
+}
+
+/* Placeholder to make future diffs smaller */
+static inline int adreno_is_a740_family(struct adreno_gpu *gpu)
+{
+ if (WARN_ON_ONCE(!gpu->info))
+ return false;
+ return gpu->info->family == ADRENO_7XX_GEN2;
+}
+
static inline int adreno_is_a7xx(struct adreno_gpu *gpu)
{
/* Update with non-fake (i.e. non-A702) Gen 7 GPUs */
- return gpu->info->family == ADRENO_7XX_GEN1;
+ return gpu->info->family == ADRENO_7XX_GEN1 ||
+ adreno_is_a740_family(gpu);
}
u64 adreno_private_address_space_size(struct msm_gpu *gpu);
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* Re: [PATCH v3 09/10] drm/msm/a6xx: Add A740 support
2023-08-23 12:56 ` [PATCH v3 09/10] drm/msm/a6xx: Add A740 support Konrad Dybcio
@ 2023-09-12 12:52 ` Konrad Dybcio
0 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-09-12 12:52 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Neil Armstrong
On 23.08.2023 14:56, Konrad Dybcio wrote:
> A740 builds upon the A730 IP, shuffling some values and registers
> around. More differences will appear when things like BCL are
> implemented.
>
> adreno_is_a740_family is added in preparation for more A7xx GPUs,
> the logic checks will be valid resulting in smaller diffs.
>
> Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
> Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
> ---
[...]
> .gmem = SZ_2M,
> .inactive_period = DRM_MSM_INACTIVE_PERIOD,
> + .quirks = ADRENO_QUIRK_HAS_CACHED_COHERENT |
> + ADRENO_QUIRK_HAS_HW_APRIV,
That's a funny conflict resolution (should have been in the previous
commit..). If there are no other comments, could you fix this up while
applying, Rob?
Konrad
^ permalink raw reply [flat|nested] 13+ messages in thread
* [PATCH v3 10/10] drm/msm/a6xx: Poll for GBIF unhalt status in hw_init
2023-08-23 12:55 [PATCH v3 00/10] A7xx support Konrad Dybcio
` (8 preceding siblings ...)
2023-08-23 12:56 ` [PATCH v3 09/10] drm/msm/a6xx: Add A740 support Konrad Dybcio
@ 2023-08-23 12:56 ` Konrad Dybcio
9 siblings, 0 replies; 13+ messages in thread
From: Konrad Dybcio @ 2023-08-23 12:56 UTC (permalink / raw)
To: Rob Clark, Abhinav Kumar, Dmitry Baryshkov, Sean Paul,
David Airlie, Daniel Vetter, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Bjorn Andersson
Cc: Marijn Suijten, linux-arm-msm, dri-devel, freedreno, devicetree,
linux-kernel, Konrad Dybcio, Neil Armstrong
Some GPUs - particularly A7xx ones - are really really stubborn and
sometimes take a longer-than-expected time to finish unhalting GBIF.
Note that this is not caused by the request a few lines above.
Poll for the unhalt ack to make sure we're not trying to write bits to
an essentially dead GPU that can't receive data on its end of the bus.
Failing to do this will result in inexplicable GMU timeouts or worse.
This is a rather ugly hack which introduces a whole lot of latency.
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # sm8450
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
index 2313620084b6..11cb410e0ac7 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
@@ -1629,6 +1629,10 @@ static int hw_init(struct msm_gpu *gpu)
mb();
}
+ /* Some GPUs are stubborn and take their sweet time to unhalt GBIF! */
+ if (adreno_is_a7xx(adreno_gpu) && a6xx_has_gbif(adreno_gpu))
+ spin_until(!gpu_read(gpu, REG_A6XX_GBIF_HALT_ACK));
+
gpu_write(gpu, REG_A6XX_RBBM_SECVID_TSB_CNTL, 0);
if (adreno_is_a619_holi(adreno_gpu))
--
2.42.0
^ permalink raw reply related [flat|nested] 13+ messages in thread