linux.git/drivers/gpu/drm/v3d, branch v6.9

drm/v3d: Don't increment `enabled_ns` twice

2024-04-15T18:00:19+00:00

The commit 509433d8146c ("drm/v3d: Expose the total GPU usage stats on sysfs")
introduced the calculation of global GPU stats. For the regards, it used
the already existing infrastructure provided by commit 09a93cc4f7d1 ("drm/v3d:
Implement show_fdinfo() callback for GPU usage stats"). While adding
global GPU stats calculation ability, the author forgot to delete the
existing one.

Currently, the value of `enabled_ns` is incremented twice by the end of
the job, when it should be added just once. Therefore, delete the
leftovers from commit 509433d8146c ("drm/v3d: Expose the total GPU usage
stats on sysfs").

Fixes: 509433d8146c ("drm/v3d: Expose the total GPU usage stats on sysfs")
Reported-by: Tvrtko Ursulin 
Signed-off-by: Maíra Canal 
Reviewed-by: Tvrtko Ursulin 
Reviewed-by: Jose Maria Casanova Crespo 
Link: https://patchwork.freedesktop.org/patch/msgid/20240403203517.731876-2-mcanal@igalia.com

drm/v3d: Enable V3D to use different PAGE_SIZE

2024-02-23T19:37:20+00:00

Currently, the V3D driver uses PAGE_SHIFT over the assumption that
PAGE_SHIFT = 12, as the PAGE_SIZE = 4KB. But, the RPi 5 is using
PAGE_SIZE = 16KB, so the MMU PAGE_SHIFT is different than the system's
PAGE_SHIFT.

Enable V3D to be used in system's with any PAGE_SIZE by making sure that
everything MMU-related uses the MMU page shift.

Signed-off-by: Maíra Canal 
Reviewed-by: Iago Toral Quiroga 
Link: https://patchwork.freedesktop.org/patch/msgid/20240214193503.164462-1-mcanal@igalia.com

Merge tag 'drm-misc-next-2024-01-11' of git://anongit.freedesktop.org/drm/drm-misc into drm-next

2024-02-05T03:50:15+00:00

drm-misc-next for v6.9:

UAPI Changes:

virtio:
- add Venus capset defines

Cross-subsystem Changes:

Core Changes:

- fix drm_fixp2int_ceil()
- documentation fixes
- clean ups
- allow DRM_MM_DEBUG with DRM=m
- build fixes for debugfs support
- EDID cleanups
- sched: error-handling fixes
- ttm: add tests

Driver Changes:

bridge:
- ite-6505: fix DP link-training bug
- samsung-dsim: fix error checking in probe
- tc358767: fix regmap usage

efifb:
- use copy of global screen_info state

hisilicon:
- fix EDID includes

mgag200:
- improve ioremap usage
- convert to struct drm_edid

nouveau:
- disp: use kmemdup()
- fix EDID includes
- documentation fixes

panel:
- ltk050h3146w: error-handling fixes
- panel-edp: support delay between power-on and enable; use put_sync in
  unprepare; support Mediatek MT8173 Chromebooks, BOE NV116WHM-N49 V8.0,
  BOE NV122WUM-N41, CSO MNC207QS1-1 plus DT bindings
- panel-lvds: support EDT ETML0700Z9NDHA plus DT bindings
- panel-novatek: FRIDA FRD400B25025-A-CTK plus DT bindings

qaic:
- fixes to BO handling
- make use of DRM managed release
- fix order of remove operations

rockchip:
- analogix_dp: get encoder port from DT
- inno_hdmi: support HDMI for RK3128
- lvds: error-handling fixes

simplefb:
- fix logging

ssd130x:
- support SSD133x plus DT bindings

tegra:
- fix error handling

tilcdc:
- make use of DRM managed release

v3d:
- show memory stats in debugfs

vc4:
- fix error handling in plane prepare_fb
- fix framebuffer test in plane helpers

vesafb:
- use copy of global screen_info state

virtio:
- cleanups

vkms:
- fix OOB access when programming the LUT
- Kconfig improvements

vmwgfx:
- unmap surface before changing plane state
- fix memory leak in error handling
- documentation fixes

Signed-off-by: Dave Airlie 

From: Thomas Zimmermann 
Link: https://patchwork.freedesktop.org/patch/msgid/20240111154902.GA8448@linux-uq9g

drm/v3d: Show the memory-management stats on debugfs

2024-01-11T14:45:47+00:00

Dump the contents of the DRM MM allocator of the V3D driver. This will
help us to debug the VA ranges allocated.

Signed-off-by: Maíra Canal 
Reviewed-by: Melissa Wen 
Link: https://patchwork.freedesktop.org/patch/msgid/20240105145851.193492-1-mcanal@igalia.com

drm/v3d: Free the job and assign it to NULL if initialization fails

2024-01-11T14:23:13+00:00

Currently, if `v3d_job_init()` fails (e.g. in the IGT test "bad-in-sync",
where we submit an invalid in-sync to the IOCTL), then we end up with
the following NULL pointer dereference:

[   34.146279] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000078
[   34.146301] Mem abort info:
[   34.146306]   ESR = 0x0000000096000005
[   34.146315]   EC = 0x25: DABT (current EL), IL = 32 bits
[   34.146322]   SET = 0, FnV = 0
[   34.146328]   EA = 0, S1PTW = 0
[   34.146334]   FSC = 0x05: level 1 translation fault
[   34.146340] Data abort info:
[   34.146345]   ISV = 0, ISS = 0x00000005, ISS2 = 0x00000000
[   34.146351]   CM = 0, WnR = 0, TnD = 0, TagAccess = 0
[   34.146357]   GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
[   34.146366] user pgtable: 4k pages, 39-bit VAs, pgdp=00000001232e6000
[   34.146375] [0000000000000078] pgd=0000000000000000, p4d=0000000000000000, pud=0000000000000000
[   34.146399] Internal error: Oops: 0000000096000005 [#1] PREEMPT SMP
[   34.146406] Modules linked in: rfcomm snd_seq_dummy snd_hrtimer snd_seq snd_seq_device algif_hash aes_neon_bs aes_neon_blk algif_skcipher af_alg bnep hid_logitech_hidpp brcmfmac_wcc brcmfmac brcmutil hci_uart vc4 btbcm cfg80211 bluetooth bcm2835_v4l2(C) snd_soc_hdmi_codec binfmt_misc cec drm_display_helper hid_logitech_dj bcm2835_mmal_vchiq(C) drm_dma_helper drm_kms_helper videobuf2_v4l2 raspberrypi_hwmon ecdh_generic videobuf2_vmalloc videobuf2_memops ecc videobuf2_common rfkill videodev libaes snd_soc_core dwc2 i2c_brcmstb snd_pcm_dmaengine snd_bcm2835(C) i2c_bcm2835 pwm_bcm2835 snd_pcm mc v3d snd_timer snd gpu_sched drm_shmem_helper nvmem_rmem uio_pdrv_genirq uio i2c_dev drm fuse dm_mod drm_panel_orientation_quirks backlight configfs ip_tables x_tables ipv6
[   34.146556] CPU: 1 PID: 1890 Comm: v3d_submit_csd Tainted: G         C         6.7.0-rc3-g49ddab089611 #68
[   34.146563] Hardware name: Raspberry Pi 4 Model B Rev 1.5 (DT)
[   34.146569] pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[   34.146575] pc : drm_sched_job_cleanup+0x3c/0x190 [gpu_sched]
[   34.146611] lr : v3d_submit_csd_ioctl+0x1b4/0x460 [v3d]
[   34.146653] sp : ffffffc083cbbb80
[   34.146658] x29: ffffffc083cbbb90 x28: ffffff81035afc00 x27: ffffffe77a641168
[   34.146668] x26: ffffff81056a8000 x25: 0000000000000058 x24: 0000000000000000
[   34.146677] x23: ffffff81065e2000 x22: ffffff81035afe00 x21: ffffffc083cbbcf0
[   34.146686] x20: ffffff81035afe00 x19: 00000000ffffffea x18: 0000000000000000
[   34.146694] x17: 0000000000000000 x16: ffffffe7989e34b0 x15: 0000000000000000
[   34.146703] x14: 0000000004000004 x13: ffffff81035afe80 x12: ffffffc083cb8000
[   34.146711] x11: cc57e05dfbe5ef00 x10: cc57e05dfbe5ef00 x9 : ffffffe77a64131c
[   34.146719] x8 : 0000000000000000 x7 : 0000000000000000 x6 : 000000000000003f
[   34.146727] x5 : 0000000000000040 x4 : ffffff81fefb03f0 x3 : ffffffc083cbba40
[   34.146736] x2 : ffffff81056a8000 x1 : ffffffe7989e35e8 x0 : 0000000000000000
[   34.146745] Call trace:
[   34.146748]  drm_sched_job_cleanup+0x3c/0x190 [gpu_sched]
[   34.146768]  v3d_submit_csd_ioctl+0x1b4/0x460 [v3d]
[   34.146791]  drm_ioctl_kernel+0xe0/0x120 [drm]
[   34.147029]  drm_ioctl+0x264/0x408 [drm]
[   34.147135]  __arm64_sys_ioctl+0x9c/0xe0
[   34.147152]  invoke_syscall+0x4c/0x118
[   34.147162]  el0_svc_common+0xb8/0xf0
[   34.147168]  do_el0_svc+0x28/0x40
[   34.147174]  el0_svc+0x38/0x88
[   34.147184]  el0t_64_sync_handler+0x84/0x100
[   34.147191]  el0t_64_sync+0x190/0x198
[   34.147201] Code: aa0003f4 f90007e8 f9401008 aa0803e0 (b8478c09)
[   34.147210] ---[ end trace 0000000000000000 ]---

This happens because we are calling `drm_sched_job_cleanup()` twice:
once at `v3d_job_init()` and again when we call `v3d_job_cleanup()`.

To mitigate this issue, we can return to the same approach that we used
to use before 464c61e76de8: deallocate the job after `v3d_job_init()`
fails and assign it to NULL. Then, when we call `v3d_job_cleanup()`, job
is NULL and the function returns.

Fixes: 464c61e76de8 ("drm/v3d: Decouple job allocation from job initiation")
Signed-off-by: Maíra Canal 
Reviewed-by: Iago Toral Quiroga 
Link: https://patchwork.freedesktop.org/patch/msgid/20240109142857.1122704-1-mcanal@igalia.com

drm/v3d: Fix support for register debugging on the RPi 4

2024-01-09T17:21:47+00:00

RPi 4 uses V3D 4.2, which is currently not supported by the register
definition stated at `v3d_core_reg_defs`. We should be able to support
V3D 4.2, therefore, change the maximum version of the register
definition to 42, not 41.

Fixes: 0ad5bc1ce463 ("drm/v3d: fix up register addresses for V3D 7.x")
Signed-off-by: Maíra Canal 
Reviewed-by: Iago Toral Quiroga 
Link: https://patchwork.freedesktop.org/patch/msgid/20240109113126.929446-1-mcanal@igalia.com

drm/v3d: Fix missing error code in v3d_submit_cpu_ioctl()

2023-12-04T22:30:33+00:00

Smatch warns:
	drivers/gpu/drm/v3d/v3d_submit.c:1222 v3d_submit_cpu_ioctl()
	warn: missing error code 'ret'

When there is no job type or job is submitted with wrong number of BOs
it is an error path, ret is zero at this point which is incorrect
return.

Fix this by changing it to -EINVAL.

Fixes: aafc1a2bea67 ("drm/v3d: Add a CPU job submission")
Signed-off-by: Harshit Mogalapalli 
Reviewed-by: Melissa Wen 
Signed-off-by: Melissa Wen 
Link: https://patchwork.freedesktop.org/patch/msgid/20231204122102.181298-1-harshit.m.mogalapalli@oracle.com

drm/v3d: Create a CPU job extension for the copy performance query job

2023-12-01T12:47:36+00:00

A CPU job is a type of job that performs operations that requires CPU
intervention. A copy performance query job is a job that copy the complete
or partial result of a query to a buffer. In order to copy the result of
a performance query to a buffer, we need to get the values from the
performance monitors.

So, create a user extension for the CPU job that enables the creation
of a copy performance query job. This user extension will allow the creation
of a CPU job that copy the results of a performance query to a BO with the
possibility to indicate the availability with a availability bit.

Signed-off-by: Maíra Canal 
Reviewed-by: Iago Toral Quiroga 
Link: https://patchwork.freedesktop.org/patch/msgid/20231130164420.932823-19-mcanal@igalia.com

drm/v3d: Create a CPU job extension for the reset performance query job

2023-12-01T12:47:35+00:00

A CPU job is a type of job that performs operations that requires CPU
intervention. A reset performance query job is a job that resets the
performance queries by resetting the values of the perfmons. Moreover,
we also reset the syncobjs related to the availability of the query.

So, create a user extension for the CPU job that enables the creation
of a reset performance job. This user extension will allow the creation of
a CPU job that resets the perfmons values and resets the availability syncobj.

Signed-off-by: Maíra Canal 
Reviewed-by: Iago Toral Quiroga 
Link: https://patchwork.freedesktop.org/patch/msgid/20231130164420.932823-18-mcanal@igalia.com

drm/v3d: Create a CPU job extension to copy timestamp query to a buffer

2023-12-01T12:47:31+00:00

A CPU job is a type of job that performs operations that requires CPU
intervention. A copy timestamp query job is a job that copy the complete
or partial result of a query to a buffer. As V3D doesn't provide any
mechanism to obtain a timestamp from the GPU, it is a job that needs
CPU intervention.

So, create a user extension for the CPU job that enables the creation
of a copy timestamp query job. This user extension will allow the creation
of a CPU job that copy the results of a timestamp query to a BO with the
possibility to indicate the timestamp availability with a availability bit.

Signed-off-by: Maíra Canal 
Reviewed-by: Iago Toral Quiroga 
Link: https://patchwork.freedesktop.org/patch/msgid/20231130164420.932823-17-mcanal@igalia.com