[v2,1/3] drm/msm/gpu: Add devfreq tuning debugfs
Commit Message
From: Rob Clark <robdclark@chromium.org>
Make the handful of tuning knobs available visible via debugfs.
v2: select DEVFREQ_GOV_SIMPLE_ONDEMAND because for some reason
struct devfreq_simple_ondemand_data depends on this
Signed-off-by: Rob Clark <robdclark@chromium.org>
---
drivers/gpu/drm/msm/Kconfig | 1 +
drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 2 +-
drivers/gpu/drm/msm/msm_debugfs.c | 12 ++++++++++++
drivers/gpu/drm/msm/msm_drv.h | 9 +++++++++
drivers/gpu/drm/msm/msm_gpu.h | 3 ---
drivers/gpu/drm/msm/msm_gpu_devfreq.c | 6 ++++--
6 files changed, 27 insertions(+), 6 deletions(-)
Comments
Hi Rob,
I love your patch! Yet something to improve:
[auto build test ERROR on drm/drm-next]
[also build test ERROR on drm-exynos/exynos-drm-next drm-intel/for-linux-next drm-intel/for-linux-next-fixes drm-tip/drm-tip linus/master v6.2-rc3 next-20230112]
[cannot apply to drm-misc/drm-misc-next]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Rob-Clark/drm-msm-gpu-Add-devfreq-tuning-debugfs/20230111-071620
base: git://anongit.freedesktop.org/drm/drm drm-next
patch link: https://lore.kernel.org/r/20230110231447.1939101-2-robdclark%40gmail.com
patch subject: [PATCH v2 1/3] drm/msm/gpu: Add devfreq tuning debugfs
config: arc-buildonly-randconfig-r003-20230112
compiler: arc-elf-gcc (GCC) 12.1.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://github.com/intel-lab-lkp/linux/commit/a5db640b1edb9bb3a67015ce8183f9b6c2e44fa0
git remote add linux-review https://github.com/intel-lab-lkp/linux
git fetch --no-tags linux-review Rob-Clark/drm-msm-gpu-Add-devfreq-tuning-debugfs/20230111-071620
git checkout a5db640b1edb9bb3a67015ce8183f9b6c2e44fa0
# save the config file
mkdir build_dir && cp config build_dir/.config
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-12.1.0 make.cross W=1 O=build_dir ARCH=arc olddefconfig
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-12.1.0 make.cross W=1 O=build_dir ARCH=arc SHELL=/bin/bash drivers/gpu/
If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@intel.com>
All errors (new ones prefixed by >>):
In file included from drivers/gpu/drm/msm/msm_kms.h:14,
from drivers/gpu/drm/msm/disp/msm_disp_snapshot.h:27,
from drivers/gpu/drm/msm/disp/msm_disp_snapshot.c:8:
>> drivers/gpu/drm/msm/msm_drv.h:238:45: error: field 'gpu_devfreq_config' has incomplete type
238 | struct devfreq_simple_ondemand_data gpu_devfreq_config;
| ^~~~~~~~~~~~~~~~~~
Kconfig warnings: (for reference only)
WARNING: unmet direct dependencies detected for DEVFREQ_GOV_SIMPLE_ONDEMAND
Depends on [n]: PM_DEVFREQ [=n]
Selected by [y]:
- DRM_MSM [=y] && HAS_IOMEM [=y] && DRM [=y] && (ARCH_QCOM || SOC_IMX5 || COMPILE_TEST [=y]) && COMMON_CLK [=y] && IOMMU_SUPPORT [=y] && (QCOM_OCMEM [=n] || QCOM_OCMEM [=n]=n) && (QCOM_LLCC [=y] || QCOM_LLCC [=y]=n) && (QCOM_COMMAND_DB [=n] || QCOM_COMMAND_DB [=n]=n)
vim +/gpu_devfreq_config +238 drivers/gpu/drm/msm/msm_drv.h
107
108 struct drm_device *dev;
109
110 struct msm_kms *kms;
111 int (*kms_init)(struct drm_device *dev);
112
113 /* subordinate devices, if present: */
114 struct platform_device *gpu_pdev;
115
116 /* possibly this should be in the kms component, but it is
117 * shared by both mdp4 and mdp5..
118 */
119 struct hdmi *hdmi;
120
121 /* DSI is shared by mdp4 and mdp5 */
122 struct msm_dsi *dsi[2];
123
124 struct msm_dp *dp[MSM_DP_CONTROLLER_COUNT];
125
126 /* when we have more than one 'msm_gpu' these need to be an array: */
127 struct msm_gpu *gpu;
128
129 /* gpu is only set on open(), but we need this info earlier */
130 bool is_a2xx;
131 bool has_cached_coherent;
132
133 struct drm_fb_helper *fbdev;
134
135 struct msm_rd_state *rd; /* debugfs to dump all submits */
136 struct msm_rd_state *hangrd; /* debugfs to dump hanging submits */
137 struct msm_perf_state *perf;
138
139 /**
140 * List of all GEM objects (mainly for debugfs, protected by obj_lock
141 * (acquire before per GEM object lock)
142 */
143 struct list_head objects;
144 struct mutex obj_lock;
145
146 /**
147 * lru:
148 *
149 * The various LRU's that a GEM object is in at various stages of
150 * it's lifetime. Objects start out in the unbacked LRU. When
151 * pinned (for scannout or permanently mapped GPU buffers, like
152 * ringbuffer, memptr, fw, etc) it moves to the pinned LRU. When
153 * unpinned, it moves into willneed or dontneed LRU depending on
154 * madvise state. When backing pages are evicted (willneed) or
155 * purged (dontneed) it moves back into the unbacked LRU.
156 *
157 * The dontneed LRU is considered by the shrinker for objects
158 * that are candidate for purging, and the willneed LRU is
159 * considered for objects that could be evicted.
160 */
161 struct {
162 /**
163 * unbacked:
164 *
165 * The LRU for GEM objects without backing pages allocated.
166 * This mostly exists so that objects are always is one
167 * LRU.
168 */
169 struct drm_gem_lru unbacked;
170
171 /**
172 * pinned:
173 *
174 * The LRU for pinned GEM objects
175 */
176 struct drm_gem_lru pinned;
177
178 /**
179 * willneed:
180 *
181 * The LRU for unpinned GEM objects which are in madvise
182 * WILLNEED state (ie. can be evicted)
183 */
184 struct drm_gem_lru willneed;
185
186 /**
187 * dontneed:
188 *
189 * The LRU for unpinned GEM objects which are in madvise
190 * DONTNEED state (ie. can be purged)
191 */
192 struct drm_gem_lru dontneed;
193
194 /**
195 * lock:
196 *
197 * Protects manipulation of all of the LRUs.
198 */
199 struct mutex lock;
200 } lru;
201
202 struct workqueue_struct *wq;
203
204 unsigned int num_crtcs;
205 struct drm_crtc *crtcs[MAX_CRTCS];
206
207 struct msm_drm_thread event_thread[MAX_CRTCS];
208
209 unsigned int num_bridges;
210 struct drm_bridge *bridges[MAX_BRIDGES];
211
212 /* VRAM carveout, used when no IOMMU: */
213 struct {
214 unsigned long size;
215 dma_addr_t paddr;
216 /* NOTE: mm managed at the page level, size is in # of pages
217 * and position mm_node->start is in # of pages:
218 */
219 struct drm_mm mm;
220 spinlock_t lock; /* Protects drm_mm node allocation/removal */
221 } vram;
222
223 struct notifier_block vmap_notifier;
224 struct shrinker shrinker;
225
226 struct drm_atomic_state *pm_state;
227
228 /**
229 * hangcheck_period: For hang detection, in ms
230 *
231 * Note that in practice, a submit/job will get at least two hangcheck
232 * periods, due to checking for progress being implemented as simply
233 * "have the CP position registers changed since last time?"
234 */
235 unsigned int hangcheck_period;
236
237 /** gpu_devfreq_config: Devfreq tuning config for the GPU. */
> 238 struct devfreq_simple_ondemand_data gpu_devfreq_config;
239
240 /**
241 * gpu_clamp_to_idle: Enable clamping to idle freq when inactive
242 */
243 bool gpu_clamp_to_idle;
244
245 /**
246 * disable_err_irq:
247 *
248 * Disable handling of GPU hw error interrupts, to force fallback to
249 * sw hangcheck timer. Written (via debugfs) by igt tests to test
250 * the sw hangcheck mechanism.
251 */
252 bool disable_err_irq;
253 };
254
On 11/01/2023 00:14, Rob Clark wrote:
> From: Rob Clark <robdclark@chromium.org>
>
> Make the handful of tuning knobs available visible via debugfs.
>
> v2: select DEVFREQ_GOV_SIMPLE_ONDEMAND because for some reason
> struct devfreq_simple_ondemand_data depends on this
>
> Signed-off-by: Rob Clark <robdclark@chromium.org>
> ---
For some reason this was merged even though earlier kbuild reported
build failure. This breaks linux next and qcom defconfig. Please drop
the patch from the linux next.
See earlier build issues reported:
https://lore.kernel.org/all/202301130108.fslQjvJ8-lkp@intel.com/
Best regards,
Krzysztof
On Mon, Jan 23, 2023 at 4:38 AM Krzysztof Kozlowski <krzk@kernel.org> wrote:
>
> On 11/01/2023 00:14, Rob Clark wrote:
> > From: Rob Clark <robdclark@chromium.org>
> >
> > Make the handful of tuning knobs available visible via debugfs.
> >
> > v2: select DEVFREQ_GOV_SIMPLE_ONDEMAND because for some reason
> > struct devfreq_simple_ondemand_data depends on this
> >
> > Signed-off-by: Rob Clark <robdclark@chromium.org>
> > ---
>
> For some reason this was merged even though earlier kbuild reported
> build failure. This breaks linux next and qcom defconfig. Please drop
> the patch from the linux next.
>
> See earlier build issues reported:
> https://lore.kernel.org/all/202301130108.fslQjvJ8-lkp@intel.com/
>
This will fix it: https://patchwork.freedesktop.org/series/113232/
BR,
-R
@@ -23,6 +23,7 @@ config DRM_MSM
select SHMEM
select TMPFS
select QCOM_SCM
+ select DEVFREQ_GOV_SIMPLE_ONDEMAND
select WANT_DEV_COREDUMP
select SND_SOC_HDMI_CODEC if SND_SOC
select SYNC_FILE
@@ -2021,7 +2021,7 @@ struct msm_gpu *a6xx_gpu_init(struct drm_device *dev)
* to cause power supply issues:
*/
if (adreno_is_a618(adreno_gpu) || adreno_is_7c3(adreno_gpu))
- gpu->clamp_to_idle = true;
+ priv->gpu_clamp_to_idle = true;
/* Check if there is a GMU phandle and set it up */
node = of_parse_phandle(pdev->dev.of_node, "qcom,gmu", 0);
@@ -305,6 +305,7 @@ void msm_debugfs_init(struct drm_minor *minor)
{
struct drm_device *dev = minor->dev;
struct msm_drm_private *priv = dev->dev_private;
+ struct dentry *gpu_devfreq;
drm_debugfs_create_files(msm_debugfs_list,
ARRAY_SIZE(msm_debugfs_list),
@@ -325,6 +326,17 @@ void msm_debugfs_init(struct drm_minor *minor)
debugfs_create_file("shrink", S_IRWXU, minor->debugfs_root,
dev, &shrink_fops);
+ gpu_devfreq = debugfs_create_dir("devfreq", minor->debugfs_root);
+
+ debugfs_create_bool("idle_clamp",0600, gpu_devfreq,
+ &priv->gpu_clamp_to_idle);
+
+ debugfs_create_u32("upthreshold",0600, gpu_devfreq,
+ &priv->gpu_devfreq_config.upthreshold);
+
+ debugfs_create_u32("downdifferential",0600, gpu_devfreq,
+ &priv->gpu_devfreq_config.downdifferential);
+
if (priv->kms && priv->kms->funcs->debugfs_init)
priv->kms->funcs->debugfs_init(priv->kms, minor);
@@ -11,6 +11,7 @@
#include <linux/kernel.h>
#include <linux/clk.h>
#include <linux/cpufreq.h>
+#include <linux/devfreq.h>
#include <linux/module.h>
#include <linux/component.h>
#include <linux/platform_device.h>
@@ -234,6 +235,14 @@ struct msm_drm_private {
*/
unsigned int hangcheck_period;
+ /** gpu_devfreq_config: Devfreq tuning config for the GPU. */
+ struct devfreq_simple_ondemand_data gpu_devfreq_config;
+
+ /**
+ * gpu_clamp_to_idle: Enable clamping to idle freq when inactive
+ */
+ bool gpu_clamp_to_idle;
+
/**
* disable_err_irq:
*
@@ -275,9 +275,6 @@ struct msm_gpu {
struct msm_gpu_state *crashstate;
- /* Enable clamping to idle freq when inactive: */
- bool clamp_to_idle;
-
/* True if the hardware supports expanded apriv (a650 and newer) */
bool hw_apriv;
@@ -183,6 +183,7 @@ static bool has_devfreq(struct msm_gpu *gpu)
void msm_devfreq_init(struct msm_gpu *gpu)
{
struct msm_gpu_devfreq *df = &gpu->devfreq;
+ struct msm_drm_private *priv = gpu->dev->dev_private;
/* We need target support to do devfreq */
if (!gpu->funcs->gpu_busy)
@@ -209,7 +210,7 @@ void msm_devfreq_init(struct msm_gpu *gpu)
df->devfreq = devm_devfreq_add_device(&gpu->pdev->dev,
&msm_devfreq_profile, DEVFREQ_GOV_SIMPLE_ONDEMAND,
- NULL);
+ &priv->gpu_devfreq_config);
if (IS_ERR(df->devfreq)) {
DRM_DEV_ERROR(&gpu->pdev->dev, "Couldn't initialize GPU devfreq\n");
@@ -358,10 +359,11 @@ static void msm_devfreq_idle_work(struct kthread_work *work)
struct msm_gpu_devfreq *df = container_of(work,
struct msm_gpu_devfreq, idle_work.work);
struct msm_gpu *gpu = container_of(df, struct msm_gpu, devfreq);
+ struct msm_drm_private *priv = gpu->dev->dev_private;
df->idle_time = ktime_get();
- if (gpu->clamp_to_idle)
+ if (priv->gpu_clamp_to_idle)
dev_pm_qos_update_request(&df->idle_freq, 0);
}