In function nvmap_ioctl_get_fd_from_list, the return pointer from nvmap_handle_get_from_id is being dereferenced without checking if it is valid. This is causing a kernel panic crash in syzkaller. Fix this by checking whether the pointer is valid or not before dereferencing it.
Bug 4479038
Change-Id: Ia65341e9eb12873e660baae44d28966e71317377
Signed-off-by: Yash Bhatt <ybhatt@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3154940
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Remove WARN_ON macro invocation and exit gracefully from the nvmap_query_heap_params function when the heap_mask parameter is not a power of two. This is because if two or more bits are set in heap_mask, then it's an invalid parameter because NvMap can allocate buffer from only one heap at a time.
Bug 4479038
Change-Id: I6cfca911115f7f29e2c4e46816a89fa1869adae4
Signed-off-by: Yash Bhatt <ybhatt@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3154939
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
As part of profiler show operation triggered through sysfs,
profiler data is only seen in the kernel log.
Add a change to copy the profiler contents in the profiler
buffer as well so that the buffer content is visible at the
place where sysfs read is issued.
Note that there is a limitation in buffer size used in sysfs
read and write operations, which is equal to PAGE_SIZE.
And so, checks are added to detect buffer overflow while
writing profiler data in the buffer.
Jira ASSA-934
Change-Id: Ice3eccdfffbd7734ac7a20a0a8ca52ba00af865e
Signed-off-by: Bharat Nihalani <bnihalani@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3154262
(cherry picked from commit b619de032dfd58ffc385b08d26353991e2ae3d9b)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3158713
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: svcacv <svcacv@nvidia.com>
Tested-by: Bitan Biswas <bbiswas@nvidia.com>
The L4T packaging needs the soc specific dummy driver to be
available always. However, it is not available in generic build and
it creates the packaging/configuration fail.
Add soc specific dummy driver for pcie so that the respective binary
should be available always.
Bug 4695516
Change-Id: Ic40e60a12de35cebc7e9acec9b09e39aa267276b
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3154314
Tested-by: Ashish Mhetre <amhetre@nvidia.com>
Reviewed-by: Bitan Biswas <bbiswas@nvidia.com>
Tested-by: Laxman Dewangan <ldewangan@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: Laxman Dewangan <ldewangan@nvidia.com>
Issue: IOCTLs from supplicanta pplication are not reaching Nvidia
Ethernet driver in kernel 6.1
Fix: Filled the correct policy while registering with netlink. Added
Cipher Suite attribute to genl macsec policy
Ported from -
https://git-master.nvidia.com/r/c/nvethernet-docs/+/3126784
Bug 4316709
Change-Id: I7d0b86d14b1d5bb9aff34b79d635bfccd7443ebf
When granule_size of type u32 and size is more than 4GB (i.e. value
having more than 32 bits), then ALIGN_GRANULE_SIZE would return
incorrect value due to ~(GRANULE_SIZE - 1), as this would be in 32
bits. Hence change the granule size's data type to size_t, so that the
above value is not restricted to 32 bits.
Bug 4650102
Change-Id: If44f2663b6a4735c193d3a80b233b53d70772335
Signed-off-by: Ketan Patil <ketanp@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3149162
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
Tested-by: Bitan Biswas <bbiswas@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
The correct module license string for drivers is "GPL" and not "GPL
V2". Note that "GPL" in the module license indicates GPLv2 according to
the kernel documentation. Using "GPL V2" causes the following kernel
warns to be printed when the module is loaded ...
r8168: module license 'GPL V2' taints kernel.
Disabling lock debugging due to kernel taint
r8168: module license taints kernel.
Fix this by correcting the license string for dummy drivers.
Bug 4606591
Bug 4471899
Change-Id: Ib785b2d3c90234e279cb1d804bf9a9842e6e5084
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3141628
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
The function blk_queue_max_hw_sectors() was removed in Linux v6.10 and
so add a test to conftest to see if this function is present or not to
allow the driver to build. The virtual-storage driver is currently
broken for Linux v5.19+ kernel due to other incompatibilities with these
newer kernels, but has been updated so that it still builds. Therefore,
allow the driver to build for Linux v6.10 while the driver is fixed
properly for these newer kernels.
Bug 4311184
Bug 4593750
Change-Id: Ic4054c87acdf3c77f03ff94fbdde513d84fe229b
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3142216
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
In Linux v6.10, the v4l2-subdev callbacks 'g_dv_timings' and
's_dv_timings' were moved from the v4l2_subdev_video_ops structure to
the v4l2_subdev_pad_ops structure. Fix the build for Linux v6.10 by
using conftest to determine which structure is used for these callbacks.
Bug 4593750
Change-Id: Ic54e88da22ed7d1da9b6026a45b9c4307637c7b4
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3142215
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Some of the conftest tests are no longer working with the mainline
kernel and this is causing various build failures for drivers. Upstream
commit a3b00f10da80 ("objpool: enable inlining objpool_push() and
objpool_pop() operations") inlined some code that causes build error to
occur when the tests are built with 'Werror=address-of-packed-member'.
Upstream commit 6f303d60534c ("gcc-9: silence 'address-of-packed-member'
warning") disables the option 'address-of-packed-member' for the kernel
already and so fix these new build failures by disabling this option
for the conftest tests.
Bug 4593750
Change-Id: I37af97079bb7734fcd5bf10384b1afc071613594
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3142214
Reviewed-by: Brad Griffis <bgriffis@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
