The buffer is also used for KMD postamble after commit 27c8851a,
so make it as big as what it would have been (PAGE_SIZE) if we
had the priviledged scratch buffer to prevent override or overflow.
Signed-off-by: Chenyang Zhong <zhongcy95@gmail.com>
Signed-off-by: Sorayukii <sorayukii69@gmail.com>
Chenyang Zhongauthored and Sorayukiicommited 2026-05-15 17:48:18 +07:00
"LA.UM.8.3.c25-09700-sdm845.0"
* tag 'LA.UM.8.3.c25-09700-sdm845.0' of https://git.codelinaro.org/clo/la/kernel/msm-4.9:
securemsm-kernel: Fixed multiple listener registration on same fd
msm: kgsl: Fix error handling during drawctxt switch
msm: kgsl: Update the protect register list
msm: kgsl: Do not release dma and anon buffers if unmap fails
soc: qcom: smem: Add boundary checks for partitions
msm: kgsl: Do not free sharedmem if it cannot be unmapped
Conflicts:
drivers/gpu/msm/kgsl_mmu.c
Change-Id: I62a1cc70492f769d35a9e91e140b1cb987269224
Currently, separate submissions are made for page table
switch and context switch to the ring buffer. However, if the
page table switch succeeds but the context switch fails,
it can lead to use of wrong page table for drawctxt.
To address this issue, rollback the pagetable to current pagetable.
Also,correctly put the refcount of adreno context during error
cleanup.
Change-Id: I1bb4ee3ebb0ce6ea32f0b6799cfb7fa89c0d09c7
Signed-off-by: Rakesh Naidu Bhaviripudi <quic_rakeshb@quicinc.com>
Rakesh Naidu Bhaviripudiauthored and SIVA MULLATIcommited 2024-07-15 17:29:56 +05:30
Update the protect register list as per the latest recommendation. Since
we are now relying on last-span-unbound feature, following are changed.
1. Last-span-unbound feature covers GMU_AO block, we have to replace
A6XX_GMU_ALWAYS_ON_COUNTER_L with A6XX_CP_ALWAYS_ON_COUNTER_LO
while accessing from GPU.
2. GMU power perf counters are unused and we are out of register address
protection spans to make them available for on older targets.
3. Remove PWR counter group safely for a6xx targets and setup the
counters internally.
4. IFPC counters are moved to part of target specific code and enabled
with ADRENO feature flag.
Change-Id: Ifcff462b669bbf9da8e4d53c87affdc8a9fcb740
Signed-off-by: Rakesh Naidu Bhaviripudi <quic_rakeshb@quicinc.com>
Rakesh Naidu Bhaviripudiauthored and Harshitha Sai Neelaticommited 2024-04-22 15:51:54 +05:30
If iommu unmap fails and leaves dma or anon buffers still mapped in the
iommu, do not free them.
Change-Id: Ice0e1a59c1ac0ee7a9d62d8899966b84fa63d5ca
Signed-off-by: Lynus Vaz <quic_lvaz@quicinc.com>
Lynus Vazauthored and Harshitha Sai Neelaticommited 2024-04-22 15:49:04 +05:30
"LA.UM.10.3.c25-02100-sdm845.0"
* tag 'LA.UM.10.3.c25-02100-sdm845.0' of https://git.codelinaro.org/clo/la/kernel/msm-4.9:
soc: qcom: smem: Add boundary checks for partitions
msm: kgsl: Update the protect register list
msm: kgsl: Do not release dma and anon buffers if unmap fails
msm: kgsl: Do not free sharedmem if it cannot be unmapped
msm: ADSPRPC: Restrict untrusted applications from attaching to GuestOS
msm: kgsl: Use dma_buf_get() to get dma_buf structure
pinctrl: qcom: Using readl_relaxed/writel_relaxed APIs
msm: adsprpc: Handle UAF in fastrpc internal munmap
iommu: Fix missing return check of arm_lpae_init_pte
bus: mhi: misc: Add check for dev_rp if it is iommu range or not
net: ipc_router: Add bounds check on tx path
msm: kgsl: Remove 'fd' dependency to get dma_buf handle
msm: kgsl: Keep postamble packets in a privileged buffer
msm: kgsl: Make sure that pool pages don't have any extra references
msm: kgsl: Prevent wrap around during user address mapping
Conflicts:
drivers/char/adsprpc.c
drivers/gpu/msm/kgsl.c
drivers/gpu/msm/kgsl_iommu.c
drivers/gpu/msm/kgsl_mmu.c
drivers/pinctrl/qcom/pinctrl-msm.c
Change-Id: I431924199616b135e180fd3a6421bb5037bcf164
Update the protect register list as per the latest recommendation. Since
we are now relying on last-span-unbound feature, following are changed.
1. Last-span-unbound feature covers GMU_AO block, we have to replace
A6XX_GMU_ALWAYS_ON_COUNTER_L with A6XX_CP_ALWAYS_ON_COUNTER_LO
while accessing from GPU.
2. GMU power perf counters are unused and we are out of register address
protection spans to make them available for on older targets.
3. Remove PWR counter group safely for a6xx targets and setup the
counters internally.
4. IFPC counters are moved to part of target specific code and enabled
with ADRENO feature flag.
Change-Id: Ifcff462b669bbf9da8e4d53c87affdc8a9fcb740
Signed-off-by: Rakesh Naidu Bhaviripudi <quic_rakeshb@quicinc.com>
If iommu unmap fails and leaves dma or anon buffers still mapped in the
iommu, do not free them.
Change-Id: Ice0e1a59c1ac0ee7a9d62d8899966b84fa63d5ca
Signed-off-by: Lynus Vaz <quic_lvaz@quicinc.com>
Lynus Vazauthored and Gayathri Veeragandamcommited 2024-02-01 14:57:31 +05:30
If sharedmem cannot be unmapped from the mmu, it can still be accessed
by the GPU. Therefore it is not safe to free the backing memory. In the
case that unmap fails, do not free it or return it to the system.
Change-Id: Iad3e86d043f129a4d71cf862865d9033d4a315e3
Signed-off-by: Lynus Vaz <quic_lvaz@quicinc.com>
Signed-off-by: Kaushal Sanadhya <quic_ksanadhy@quicinc.com>
Kaushal Sanadhyaauthored and Harshitha Sai Neelaticommited 2023-12-29 18:29:23 +05:30
If sharedmem cannot be unmapped from the mmu, it can still be accessed
by the GPU. Therefore it is not safe to free the backing memory. In the
case that unmap fails, do not free it or return it to the system.
Change-Id: Iad3e86d043f129a4d71cf862865d9033d4a315e3
Signed-off-by: Lynus Vaz <quic_lvaz@quicinc.com>
Signed-off-by: Kaushal Sanadhya <quic_ksanadhy@quicinc.com>
Kaushal Sanadhyaauthored and Rakesh Naidu Bhaviripudicommited 2023-12-22 12:56:00 +05:30
When setting svm region during the gpuobj import ioctl call for a usermem
address, there is a possibility of a very large input size causing the
region's 64-bit end address to wrap around. This can cause the region
to incorrectly be considered valid, ultimately allowing a use after free
scenario. To prevent this, detect the occurrence of a wrap and reject the
import.
Change-Id: I4a88f56c58b830d4342e47dc1d1f6290c78ab6b4
Signed-off-by: Mohammed Mirza Mandayappurath Manzoor <quic_mmandaya@quicinc.com>
Signed-off-by: Puranam V G Tejaswi <quic_pvgtejas@quicinc.com>
Mohammed Mirza Mandayappurath Manzoorauthored and Michael Bestascommited 2023-12-12 01:48:34 +02:00
Currently we don't ensure if vma->vm_file is associated with dma_buf. This
can cause issues later when private_data from a non dma_buf file is used as
dma_buf structure. Hence get the fd that is associated with vma->vm_file
and use dma_buf_get() to get pointer to dma_buf structure. dma_buf_get()
ensures that the file from the input fd is associated with dma_buf.
Change-Id: Ib78aef8b16bedca5ca86d3a132278ff9f07dce73
Signed-off-by: Puranam V G Tejaswi <quic_pvgtejas@quicinc.com>
Puranam V G Tejaswiauthored and Rakesh Naidu Bhaviripudicommited 2023-11-28 10:34:02 +05:30
Get the dma_buf handle directly from 'vm_file' after
doing necessary checks on the file.
Change-Id: Id5eec16588d64e4e28483b32bb52d4d3d9b86b99
Signed-off-by: ravnar <quic_ravnar@quicinc.com>
Signed-off-by: Sanjay Yadav <quic_sanjyada@quicinc.com>
Signed-off-by: Abhishek Barman <quic_abarman@quicinc.com>
Signed-off-by: Kamal Agrawal <quic_kamaagra@quicinc.com>
Signed-off-by: NISARG SHETH <quic_nsheth@quicinc.com>
Postamble packets are executed in privileged mode by gpu. So we should keep
them in a privileged scratch buffer to block userspace access. For
targets with APRIV feature support, we can mark the preemption scratch
buffer as privileged too to avoid similar issues in future.
Change-Id: Ifda360dda251083f38dfde80ce1b5dc83daae902
Signed-off-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Signed-off-by: Kaushal Sanadhya <quic_ksanadhy@quicinc.com>
Signed-off-by: NISARG SHETH <quic_nsheth@quicinc.com>
Before putting a page back in the pool be sure that it doesn't have
any additional references that would be a signal that somebody else
is looking at the page and that it would be a bad idea to keep it
around and run the risk of accidentally handing it to a different
process.
Change-Id: Ic0dedbad0cf2ffb34b76ad23e393c5a911114b82
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Harshitha Sai Neelati <quic_hsaineel@quicinc.com>
Signed-off-by: NISARG SHETH <quic_nsheth@quicinc.com>
When setting svm region during the gpuobj import ioctl call for a usermem
address, there is a possibility of a very large input size causing the
region's 64-bit end address to wrap around. This can cause the region
to incorrectly be considered valid, ultimately allowing a use after free
scenario. To prevent this, detect the occurrence of a wrap and reject the
import.
Change-Id: I4a88f56c58b830d4342e47dc1d1f6290c78ab6b4
Signed-off-by: Mohammed Mirza Mandayappurath Manzoor <quic_mmandaya@quicinc.com>
Signed-off-by: NISARG SHETH <quic_nsheth@quicinc.com>
When setting svm region during the gpuobj import ioctl call for a usermem
address, there is a possibility of a very large input size causing the
region's 64-bit end address to wrap around. This can cause the region
to incorrectly be considered valid, ultimately allowing a use after free
scenario. To prevent this, detect the occurrence of a wrap and reject the
import.
Change-Id: I4a88f56c58b830d4342e47dc1d1f6290c78ab6b4
Signed-off-by: Mohammed Mirza Mandayappurath Manzoor <quic_mmandaya@quicinc.com>
Mohammed Mirza Mandayappurath Manzoorauthored and NISARG SHETHcommited 2023-09-21 16:38:01 +05:30
"LA.UM.8.3.c25-08000-sdm845.0"
* tag 'LA.UM.8.3.c25-08000-sdm845.0' of https://git.codelinaro.org/clo/la/kernel/msm-4.9:
msm: kgsl: Keep postamble packets in a privileged buffer
msm: kgsl: Make sure that pool pages don't have any extra references
msm: kgsl: Use dma_buf_get() to get dma_buf structure
net: ipc_router: Add bounds check on tx path
Conflicts:
drivers/gpu/msm/kgsl.c
Change-Id: Icf8c4ad9804ab5636b8a10b85e7cfc641f82b6d2
Before putting a page back in the pool be sure that it doesn't have
any additional references that would be a signal that somebody else
is looking at the page and that it would be a bad idea to keep it
around and run the risk of accidentally handing it to a different
process.
Change-Id: Ic0dedbad0cf2ffb34b76ad23e393c5a911114b82
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Kamal Agrawal <quic_kamaagra@quicinc.com>
Kamal Agrawalauthored and Michael Bestascommited 2023-05-02 20:36:07 +03:00
Currently we don't ensure if vma->vm_file is associated with dma_buf. This
can cause issues later when private_data from a non dma_buf file is used as
dma_buf structure. Hence get the fd that is associated with vma->vm_file
and use dma_buf_get() to get pointer to dma_buf structure. dma_buf_get()
ensures that the file from the input fd is associated with dma_buf.
Change-Id: Ib78aef8b16bedca5ca86d3a132278ff9f07dce73
Signed-off-by: Puranam V G Tejaswi <quic_pvgtejas@quicinc.com>
Puranam V G Tejaswiauthored and Michael Bestascommited 2023-05-02 20:36:07 +03:00
Postamble packets are executed in privileged mode by gpu. So we should keep
them in a privileged scratch buffer to block userspace access. For
targets with APRIV feature support, we can mark the preemption scratch
buffer as privileged too to avoid similar issues in future.
Change-Id: Ifda360dda251083f38dfde80ce1b5dc83daae902
Signed-off-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Signed-off-by: Kaushal Sanadhya <quic_ksanadhy@quicinc.com>
Akhil P Oommenauthored and Gerrit - the friendly Code Review servercommited 2023-04-18 23:22:17 -07:00
Before putting a page back in the pool be sure that it doesn't have
any additional references that would be a signal that somebody else
is looking at the page and that it would be a bad idea to keep it
around and run the risk of accidentally handing it to a different
process.
Change-Id: Ic0dedbad0cf2ffb34b76ad23e393c5a911114b82
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Harshitha Sai Neelati <quic_hsaineel@quicinc.com>
Harshitha Sai Neelatiauthored and Rajesh Kemisetticommited 2023-03-05 20:20:35 +05:30
Currently we don't ensure if vma->vm_file is associated with dma_buf. This
can cause issues later when private_data from a non dma_buf file is used as
dma_buf structure. Hence get the fd that is associated with vma->vm_file
and use dma_buf_get() to get pointer to dma_buf structure. dma_buf_get()
ensures that the file from the input fd is associated with dma_buf.
Change-Id: Ib78aef8b16bedca5ca86d3a132278ff9f07dce73
Signed-off-by: Puranam V G Tejaswi <quic_pvgtejas@quicinc.com>
Puranam V G Tejaswiauthored and Rajesh Kemisetticommited 2023-03-05 08:43:08 +05:30
"LA.UM.8.3.c25-07000-sdm845.0"
* tag 'LA.UM.8.3.c25-07000-sdm845.0' of https://git.codelinaro.org/clo/la/kernel/msm-4.9:
msm: ADSPRPC: Restrict untrusted applications from attaching to GuestOS
msm: kgsl: Remove 'fd' dependency to get dma_buf handle
msm: kgsl: Fix gpuaddr_in_range() to check upper bound
msm: kgsl: Zap performance counters across context switches
msm: kgsl: Add a sysfs node to control performance counter reads
msm: kgsl: Perform cache flush on the pages obtained using get_user_pages()
msm: kgsl: Fix memory leak for anonymous buffers
msm: kgsl: Update register protection config
disp: msm: sde: protect file private structure with mutex lock
msm: kgsl: Fix out of bound write in adreno_profile_submit_time
soc: qcom: glink: wait for tx blocked signal response
msm: kgsl: Deregister gpu address on memdesc_sg_virt failure
disp: msm: sde: add null check for drm file in msm_release
msm: adsprpc: Handle UAF in process shell memory
msm: adsprpc: Fix race condition in internal_control
msm: adsprpc: overflow vulnerability by race condition in adsprpc driver
crypto: Fix possible stack out-of-bound error
net:sockev: hold file reference till the sock event is sent
msm: kgsl: Access map_count only if entry is successfully allocated
msm: kgsl: Correct the refcount on current process PID
msm: kgsl: Stop using memdesc->usermem
msm: kgsl: Protect the memdesc->gpuaddr in SVM use cases
msm:ADSPRPC :Fix to avoid Use after free in fastrpc_internal_munmap
msm: kgsl: Allocate user_ctxt_record based on preemption feature flag
Conflicts:
drivers/char/adsprpc.c
drivers/gpu/msm/kgsl_iommu.c
drivers/gpu/msm/kgsl_mmu.c
Change-Id: I918030d8b29f3a67939a7711b4fa9f65da4c02b8
Get the dma_buf handle directly from 'vm_file' after
doing necessary checks on the file.
Change-Id: Id5eec16588d64e4e28483b32bb52d4d3d9b86b99
Signed-off-by: ravnar <quic_ravnar@quicinc.com>
Signed-off-by: Sanjay Yadav <quic_sanjyada@quicinc.com>
Signed-off-by: Abhishek Barman <quic_abarman@quicinc.com>
Signed-off-by: Kamal Agrawal <quic_kamaagra@quicinc.com>
Abhishek Barmanauthored and Kamal Agrawalcommited 2022-05-11 15:07:20 +05:30
Currently gpuaddr_in_range() accepts only the gpuaddr & returns
true if it lies in valid range. But this does not mean that the
entire buffer is within range.
Modify the function to accept size as a parameter and check that
both starting & ending points of buffer lie within mmu range.
Change-Id: I1d722295b9a27e746bfdb6d3bf409ffe722193cb
Signed-off-by: Rohan Sethi <quic_rohsethi@quicinc.com>
Signed-off-by: Abhishek Barman <quic_abarman@quicinc.com>
Signed-off-by: Kamal Agrawal <quic_kamaagra@quicinc.com>
Abhishek Barmanauthored and Kamal Agrawalcommited 2022-05-11 14:59:18 +05:30
"LA.UM.10.3.r1-01000-sdm845.0"
* tag 'LA.UM.10.3.r1-01000-sdm845.0' of https://source.codeaurora.org/quic/la/kernel/msm-4.9:
msm: camera: sensor: Add copyright details
msm: adsprpc: Handle UAF in fastrpc debugfs read
Input: uinput - avoid FF flush when destroying device
drm/dp: via debugfs entry node perform power transition
msm: kgsl: Perform cache flush on the pages obtained using get_user_pages()
drm/atomic: Remove WARN_ON for invalid plane configuration.
msm: camera: sensor: Add support for sensor read
drm/dp: handle power state transitions
scripts: checkpatch: Allow quicinc.com identity
arm: dts: msm: regulator change to power up imx334 lc sensor on hp device
msm: camera: isp: Fix race condition b/w add and apply req
ion: Fix integer overflow in msm_ion_custom_ioctl
ipa: Null persistent pointers after free
drm/msm/dp: read extended dpcd if present
Makefile: Fix GCC_TOOLCHAIN_DIR prefix for Clang cross compilation
msm: camera: jpeg: Ensure in/out map entries are within allowed range
msm-ipa: fix use-after-free of rt_tbl
diag: Prevent out of bound write while sending dci pkt to remote
Change-Id: I12e95a812e3aca20abaa10ab2c89e453eadb1769
Performance counter values need not be retained across contexts unless
specifically requested for debug. Zap the counters by initialising
perfcounter SRAM with 0's using GPU_RBBM_PERFCTR_SRAM_INIT_CMD.
Add pm4 packets during context switches and add a KMD postamble packet to
clear the counters during preemption. Do not enable perfcounter save and
restore unless requested.
Change-Id: I371779ce659c07a1cc664327f5ecdcf0374201d8
Signed-off-by: Mohammed Mirza Mandayappurath Manzoor <quic_mmandaya@quicinc.com>
Signed-off-by: Sebanti Das <quic_sebadas@quicinc.com>
Signed-off-by: Pankaj Gupta <quic_gpankaj@quicinc.com>
Signed-off-by: Harshitha Sai Neelati <quic_hsaineel@quicinc.com>
Pankaj Guptaauthored and Gerrit - the friendly Code Review servercommited 2022-03-23 05:03:19 -07:00
Currently performance counters are global and can be read by anyone. Change
the behaviour to disable reading global counters as default and add a sysfs
node to enable/disable reads.
Change-Id: Ic3785acd9bd7425c2a844ed103d7b870d9f80adf
Signed-off-by: Mohammed Mirza Mandayappurath Manzoor <quic_mmandaya@quicinc.com>
Signed-off-by: Pankaj Gupta <quic_gpankaj@quicinc.com>
Signed-off-by: Harshitha Sai Neelati <quic_hsaineel@quicinc.com>
Pankaj Guptaauthored and Gerrit - the friendly Code Review servercommited 2022-03-23 05:02:35 -07:00
Performance counter values need not be retained across contexts unless
specifically requested for debug. Zap the counters by initialising
perfcounter SRAM with 0's using GPU_RBBM_PERFCTR_SRAM_INIT_CMD.
Add pm4 packets during context switches and add a KMD postamble packet to
clear the counters during preemption. Do not enable perfcounter save and
restore unless requested.
Change-Id: I371779ce659c07a1cc664327f5ecdcf0374201d8
Signed-off-by: Mohammed Mirza Mandayappurath Manzoor <quic_mmandaya@quicinc.com>
Signed-off-by: Sebanti Das <quic_sebadas@quicinc.com>
Signed-off-by: Pankaj Gupta <quic_gpankaj@quicinc.com>
Pankaj Guptaauthored and Sebanti Dascommited 2022-03-21 13:36:41 +05:30
Currently performance counters are global and can be read by anyone. Change
the behaviour to disable reading global counters as default and add a sysfs
node to enable/disable reads.
Change-Id: Ic3785acd9bd7425c2a844ed103d7b870d9f80adf
Signed-off-by: Mohammed Mirza Mandayappurath Manzoor <quic_mmandaya@quicinc.com>
Signed-off-by: Harshitha Sai Neelati <quic_hsaineel@quicinc.com>
Signed-off-by: Pankaj Gupta <quic_gpankaj@quicinc.com>
Signed-off-by: Sebanti Das <quic_sebadas@quicinc.com>
Pankaj Guptaauthored and Sebanti Dascommited 2022-03-21 13:32:45 +05:30
Consider a scenario where user allocates anonymous memory but does not
write to it. Here the physical pages are not yet allocated. Now when this
memory is requested to be imported, a list of newly allocated zero pages
is obtained using get_user_pages(). Currently cache flush is not done for
these pages and hence GPU sees stale data. Fix this by performing cache
flush on these pages.
Change-Id: Id1e8aa20e8a9de112761732ed92f30c01088840b
Signed-off-by: Puranam V G Tejaswi <quic_pvgtejas@quicinc.com>
Signed-off-by: Sebanti Das <quic_sebadas@quicinc.com>
Signed-off-by: Kamal Agrawal <quic_kamaagra@quicinc.com>
Puranam V G Tejaswiauthored and Kamal Agrawalcommited 2022-02-01 12:44:29 +05:30
Consider a scenario where user allocates anonymous memory but does not
write to it. Here the physical pages are not yet allocated. Now when this
memory is requested to be imported, a list of newly allocated zero pages
is obtained using get_user_pages(). Currently cache flush is not done for
these pages and hence GPU sees stale data. Fix this by performing cache
flush on these pages.
Change-Id: Id1e8aa20e8a9de112761732ed92f30c01088840b
Signed-off-by: Puranam V G Tejaswi <quic_pvgtejas@quicinc.com>
Signed-off-by: Sebanti Das <quic_sebadas@quicinc.com>
Puranam V G Tejaswiauthored and Gerrit - the friendly Code Review servercommited 2022-01-19 23:35:45 -08:00
"LA.UM.10.3.r1-00500-sdm845.0"
* tag 'LA.UM.10.3.r1-00500-sdm845.0' of https://source.codeaurora.org/quic/la/kernel/msm-4.9:
tty: Fix ->pgrp locking in tiocspgrp()
msm: kgsl: Update register protection config
msm: camera: req_mgr: Table info dump removed
msm: camera: cdm: Acquire mutex lock before accessing client data
disp: msm: sde: protect file private structure with mutex lock
ARM: dts: msm: update subtype for qcs605 LC ipcamera
msm: camera: jpeg: Ensure in/out map entries are within allowed range
defconfig: enable splash logo via defconfig
drm: msm: dp: prevent calling client_register multiple times
drm: read boolan flag to enable static splash
drm: drm bootsplash enable for display port
drm: add single commit in place of multiple commits
drm: call drm client for bootsplash entry
drm/client: Hack: Add bootsplash example
drm: bitmap image of splash logo
drm: include function pointer for vmap
drm: Add drm_connector_for_each_possible_encoder()
drm/modes: Introduce drm_mode_match()
drm/file: Don't set master on in-kernel clients
drm/debugfs: Fix framebuffer debugfs file init
drm/client: Support unmapping of DRM client buffers
drm/modes: Skip invalid cmdline mode
msm: kgsl: Fix out of bound write in adreno_profile_submit_time
drm/modes: Support modes names on the command line
drm/client: Restrict the plane_state scope
drm/fb-helper: Move out modeset config code
drm: Add drm_connector_for_each_possible_encoder()
drm/modes: Introduce drm_mode_match()
drm/fb-helper: Move out commit code
drm/fb-helper: Avoid race with DRM userspace
drm: move drv test macros out of drmP.h
drm/atomic: Add macros to access existing old/new state, v2.
drm/atomic: Add new iterators over all state, v3.
drm: Add DRM_MODE_ROTATE_ and DRM_MODE_REFLECT_ to UAPI
drm/fb-helper: Remove drm_fb_helper_crtc
drm/client: Rename drm_client_add() to drm_client_register()
drm/cma-helper: Fix crash in fbdev error path
drm/client: Fix double free in error path
drm/client: Fix: drm_client_new: Don't require DRM to be registered
drm/debugfs: Add internal client debugfs file
drm: Begin an API for in-kernel clients
drm/framebuffer: Print task that allocated the fb in debug info.
drm: Make ioctls available for in-kernel clients
drm: provide management functions for drm_file
uapi: sound: add bt external sink delay parameter
msm: kgsl: Fix memory leak for anonymous buffers
uapi: Add UAPI headers for slatecom_interface driver
drm/msm/sde: enable continuous splash for external bridge
drm/msm/dsi-staging: skip regulator configuration for external bridge
diag: Add protection while accessing diag client map
mhi: core: Fix find_last_bit() usage
ARM: dts: msm: add bootloader enabled regulators for QCS605 MTP
Revert "usb: gadget: Add check gadget function bind or not"
Revert "usb: gadget: f_ecm: allocate/free net device upon driver bind/unbind"
Revert "usb: gadget: f_rndis: allocate/free net device upon driver bind/unbind"
Revert "usb: gadget: f_ncm: allocate/free net device upon driver bind/unbind"
diag: Use file's private data to get pid of exiting process
drm/msm/dsi-staging: fix dsi clock calculation for low bpp
drivers/clk/qcom: Remove 2/3 divider from frac_table_pixel
ARM: dts: msm: correct cdc reset contol active voltage level
msm: vidc: Enable dual core in LP for higher resolutions
ARM: dts: Remove disabled camera nodes from sdxpoorwills
Use environment variable to find unifdef tool
Change the subprocess argument to fix the build issue
kernel_headers: Explicitly run headers_install under 'sh'
disp: msm: sde: add null check for drm file in msm_release
msm: adsprpc: Handle UAF in process shell memory
ARM: dts: msm: Update CDSP pil region
msm: ADSPRPC: Fix return value when dsp capabilities is not supported
Conflicts:
drivers/tty/tty_io.c
Change-Id: Icfcd7b69025e4b6475416a518fa64f14ca89bc54
Currently, clean up is not done properly for anonymous buffer
(KGSL_MEM_ENTRY_USER). Fix it by freeing up resources allocated
during memdesc_sg_virt.
Change-Id: I75bff2e718b494c102a8075a5f27323e1823b212
Signed-off-by: Kamal Agrawal <kamaagra@codeaurora.org>
Signed-off-by: Rajesh Kemisetti <rajeshk@codeaurora.org>
Kamal Agrawalauthored and Rajesh Kemisetticommited 2021-12-23 18:45:24 +05:30
Update the register protection configuration as per the latest
recommendation.
Change-Id: I70365268e8e4c7ee4ff28538f22e970822e4edf0
Signed-off-by: Akhil P Oommen <akhilpo@codeaurora.org>
Akhil P Oommenauthored and Kamal Agrawalcommited 2021-11-08 12:21:41 +05:30
Update the register protection configuration as per the latest
recommendation.
Change-Id: I70365268e8e4c7ee4ff28538f22e970822e4edf0
Signed-off-by: Akhil P Oommen <akhilpo@codeaurora.org>
Akhil P Oommenauthored and Kamal Agrawalcommited 2021-11-08 12:17:13 +05:30
Make sure there is enough room in the memory descriptor to store the
entire profiling buffer object.
Change-Id: I1e1c73097bb2bba9645b0a3c66fdbbc71d8ba8fa
Signed-off-by: Kamal Agrawal <kamaagra@codeaurora.org>
Kamal Agrawalauthored and Rajesh Kemisetticommited 2021-09-28 11:52:14 +05:30
Make sure there is enough room in the memory descriptor to store the
entire profiling buffer object.
Change-Id: I1e1c73097bb2bba9645b0a3c66fdbbc71d8ba8fa
Signed-off-by: Kamal Agrawal <kamaagra@codeaurora.org>
Kamal Agrawalauthored and Gerrit - the friendly Code Review servercommited 2021-09-27 07:23:22 -07:00
Currently, clean up is not done properly for anonymous buffer
(KGSL_MEM_ENTRY_USER). Fix it by freeing up resources allocated
during memdesc_sg_virt.
Change-Id: I75bff2e718b494c102a8075a5f27323e1823b212
Signed-off-by: Kamal Agrawal <kamaagra@codeaurora.org>
Signed-off-by: Pranav Patel <pranavp@codeaurora.org>
Kamal Agrawalauthored and Pranav Patelcommited 2021-09-16 17:59:40 +05:30
When memdesc_sg_virt returns error in kgsl_setup_anon_useraddr function,
the gpu address registered in SVM region will not be deregistered.
This change deregisters the gpu address on memdesc_sg_virt failure.
Change-Id: Ic99167e283a0c6331bb9f5f0b608b6cdb3c918e4
Signed-off-by: Neeraja P <neerp@codeaurora.org>
Signed-off-by: Rajesh Kemisetti <rajeshk@codeaurora.org>
Rajesh Kemisettiauthored and Harshitha Sai Neelaticommited 2021-08-16 13:57:05 +05:30
do_shrink_slab() scans each shrinker in batches of at most
batch_size (128) pages at a time until total_scan pages are
scanned or until shrinker returns SHRINK_STOP. Under heavy
memory pressure total_scan can be large (in thousands) and
kgsl_pool_shrink_scan_objects() ends up returning 0 after
all pages that were reclaimable are reclaimed. This results in
multiple calls to kgsl_pool_shrink_scan_objects() that do not
reclaim any memory. To prevent this kgsl_pool_shrink_scan_objects()
is modified to return SHRINK_STOP as soon as no more memory can
be reclaimed.
Bug: 69931996
Test: tested using alloc-stress with additional traces
Change-Id: Ia48fc2c0d888c54ec9642c0b0962a70ca3cb4c5e
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Suren Baghdasaryanauthored and Michael Bestascommited 2021-05-26 19:41:12 +03:00
Chenyang Zhong
authored and
Sorayukii
commited 
Michael Bestas
commited 
Rakesh Naidu Bhaviripudi
authored and
SIVA MULLATI
commited 
Harshitha Sai Neelati
commited 
Lynus Vaz
authored and
Gayathri Veeragandam
commited 
Kaushal Sanadhya
authored and
Mohammed Mirza Mandayappurath Manzoor
authored and
Puranam V G Tejaswi
authored and
NISARG SHETH
commited 
Kamal Agrawal
authored and
Akhil P Oommen
authored and
Gerrit - the friendly Code Review server
commited 
Linux Build Service Account
authored and
Rajesh Kemisetti
commited 
Abhishek Barman
authored and
Pankaj Gupta
authored and
Sebanti Das
commited 
Kamal Agrawal
authored and
Akhil P Oommen
authored and
Rajesh Kemisetti
commited 
Pranav Patel
commited 
Harshitha Sai Neelati
commited 
Suren Baghdasaryan
authored and