kernel/linux.git/drivers/tee/tee_shm.c, branch v6.6.148

tee: fix NULL pointer dereference in tee_shm_put

2025-09-09T16:56:20+00:00

[ Upstream commit e4a718a3a47e89805c3be9d46a84de1949a98d5d ] tee_shm_put have NULL pointer dereference: __optee_disable_shm_cache --> shm = reg_pair_to_ptr(...);//shm maybe return NULL tee_shm_free(shm); --> tee_shm_put(shm);//crash Add check in tee_shm_put to fix it. panic log: Unable to handle kernel paging request at virtual address 0000000000100cca Mem abort info: ESR = 0x0000000096000004 EC = 0x25: DABT (current EL), IL = 32 bits SET = 0, FnV = 0 EA = 0, S1PTW = 0 FSC = 0x04: level 0 translation fault Data abort info: ISV = 0, ISS = 0x00000004, ISS2 = 0x00000000 CM = 0, WnR = 0, TnD = 0, TagAccess = 0 GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0 user pgtable: 4k pages, 48-bit VAs, pgdp=0000002049d07000 [0000000000100cca] pgd=0000000000000000, p4d=0000000000000000 Internal error: Oops: 0000000096000004 [#1] SMP CPU: 2 PID: 14442 Comm: systemd-sleep Tainted: P OE ------- ---- 6.6.0-39-generic #38 Source Version: 938b255f6cb8817c95b0dd5c8c2944acfce94b07 Hardware name: greatwall GW-001Y1A-FTH, BIOS Great Wall BIOS V3.0 10/26/2022 pstate: 80000005 (Nzcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--) pc : tee_shm_put+0x24/0x188 lr : tee_shm_free+0x14/0x28 sp : ffff001f98f9faf0 x29: ffff001f98f9faf0 x28: ffff0020df543cc0 x27: 0000000000000000 x26: ffff001f811344a0 x25: ffff8000818dac00 x24: ffff800082d8d048 x23: ffff001f850fcd18 x22: 0000000000000001 x21: ffff001f98f9fb88 x20: ffff001f83e76218 x19: ffff001f83e761e0 x18: 000000000000ffff x17: 303a30303a303030 x16: 0000000000000000 x15: 0000000000000003 x14: 0000000000000001 x13: 0000000000000000 x12: 0101010101010101 x11: 0000000000000001 x10: 0000000000000001 x9 : ffff800080e08d0c x8 : ffff001f98f9fb88 x7 : 0000000000000000 x6 : 0000000000000000 x5 : 0000000000000000 x4 : 0000000000000000 x3 : 0000000000000000 x2 : ffff001f83e761e0 x1 : 00000000ffff001f x0 : 0000000000100cca Call trace: tee_shm_put+0x24/0x188 tee_shm_free+0x14/0x28 __optee_disable_shm_cache+0xa8/0x108 optee_shutdown+0x28/0x38 platform_shutdown+0x28/0x40 device_shutdown+0x144/0x2b0 kernel_power_off+0x3c/0x80 hibernate+0x35c/0x388 state_store+0x64/0x80 kobj_attr_store+0x14/0x28 sysfs_kf_write+0x48/0x60 kernfs_fop_write_iter+0x128/0x1c0 vfs_write+0x270/0x370 ksys_write+0x6c/0x100 __arm64_sys_write+0x20/0x30 invoke_syscall+0x4c/0x120 el0_svc_common.constprop.0+0x44/0xf0 do_el0_svc+0x24/0x38 el0_svc+0x24/0x88 el0t_64_sync_handler+0x134/0x150 el0t_64_sync+0x14c/0x15 Fixes: dfd0743f1d9e ("tee: handle lookup of shm with reference count 0") Signed-off-by: Pei Xiao Reviewed-by: Sumit Garg Signed-off-by: Jens Wiklander Signed-off-by: Sasha Levin

tee: Pass a pointer to virt_to_page()

2023-03-30T06:34:52+00:00

Like the other calls in this function virt_to_page() expects a pointer, not an integer. However since many architectures implement virt_to_pfn() as a macro, this function becomes polymorphic and accepts both a (unsigned long) and a (void *). Fix this up with an explicit cast. Signed-off-by: Linus Walleij Signed-off-by: Jens Wiklander

tee: Remove call to get_kernel_pages()

2023-02-13T13:16:40+00:00

The kernel pages used by shm_get_kernel_pages() are allocated using GFP_KERNEL through the following call stack: trusted_instantiate() trusted_payload_alloc() -> GFP_KERNEL tee_shm_register_kernel_buf() register_shm_helper() shm_get_kernel_pages() Where is one of: trusted_key_unseal() trusted_key_get_random() trusted_key_seal() Because the pages can't be from highmem get_kernel_pages() boils down to a get_page() call. Remove the get_kernel_pages() call and open code the get_page(). In case a highmem page does slip through warn on once for a kmap'ed address. Cc: Jens Wiklander Cc: Al Viro Cc: "Fabio M. De Francesco" Cc: Christoph Hellwig Cc: Linus Torvalds Signed-off-by: Ira Weiny Reviewed-by: Christoph Hellwig Reviewed-by: Sumit Garg Signed-off-by: Jens Wiklander

tee: Remove vmalloc page support

2023-02-13T13:16:39+00:00

The kernel pages used by shm_get_kernel_pages() are allocated using GFP_KERNEL through the following call stack: trusted_instantiate() trusted_payload_alloc() -> GFP_KERNEL tee_shm_register_kernel_buf() register_shm_helper() shm_get_kernel_pages() Where is one of: trusted_key_unseal() trusted_key_get_random() trusted_key_seal() Remove the vmalloc page support from shm_get_kernel_pages(). Replace with a warn on once. Cc: Al Viro Cc: "Fabio M. De Francesco" Cc: Christoph Hellwig Cc: Linus Torvalds Reviewed-by: Jens Wiklander Signed-off-by: Ira Weiny Reviewed-by: Christoph Hellwig Reviewed-by: Sumit Garg Signed-off-by: Jens Wiklander

tee: fix compiler warning in tee_shm_register()

2022-08-25T09:40:06+00:00

Include to avoid the warning: drivers/tee/tee_shm.c: In function 'tee_shm_register': >> drivers/tee/tee_shm.c:242:14: error: implicit declaration of function 'access_ok' [-Werror=implicit-function-declaration] 242 | if (!access_ok((void __user *)addr, length)) | ^~~~~~~~~ cc1: some warnings being treated as errors Fixes: 573ae4f13f63 ("tee: add overflow check in register_shm_helper()") Reviewed-by: Sumit Garg Reported-by: kernel test robot Signed-off-by: Jens Wiklander

tee: add overflow check in register_shm_helper()

2022-08-18T16:20:27+00:00

With special lengths supplied by user space, register_shm_helper() has an integer overflow when calculating the number of pages covered by a supplied user space memory region. This causes internal_get_user_pages_fast() a helper function of pin_user_pages_fast() to do a NULL pointer dereference: Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 Modules linked in: CPU: 1 PID: 173 Comm: optee_example_a Not tainted 5.19.0 #11 Hardware name: QEMU QEMU Virtual Machine, BIOS 0.0.0 02/06/2015 pc : internal_get_user_pages_fast+0x474/0xa80 Call trace: internal_get_user_pages_fast+0x474/0xa80 pin_user_pages_fast+0x24/0x4c register_shm_helper+0x194/0x330 tee_shm_register_user_buf+0x78/0x120 tee_ioctl+0xd0/0x11a0 __arm64_sys_ioctl+0xa8/0xec invoke_syscall+0x48/0x114 Fix this by adding an an explicit call to access_ok() in tee_shm_register_user_buf() to catch an invalid user space address early. Fixes: 033ddf12bcf5 ("tee: add register user memory") Cc: stable@vger.kernel.org Reported-by: Nimish Mishra Reported-by: Anirban Chakraborty Reported-by: Debdeep Mukhopadhyay Suggested-by: Jerome Forissier Signed-off-by: Jens Wiklander Signed-off-by: Linus Torvalds

Merge tag 'tee-cleanup-for-v5.19' of https://git.linaro.org/people/jens.wiklander/linux-tee into arm/drivers

2022-05-06T20:02:12+00:00

TEE cleanup Removes the old and unused TEE_IOCTL_SHM_* flags Removes unused the unused tee_shm_va2pa() and tee_shm_pa2va() functions * tag 'tee-cleanup-for-v5.19' of https://git.linaro.org/people/jens.wiklander/linux-tee: tee: remove flags TEE_IOCTL_SHM_MAPPED and TEE_IOCTL_SHM_DMA_BUF tee: remove tee_shm_va2pa() and tee_shm_pa2va() Link: https://lore.kernel.org/r/20220506070328.GA1344495@jade Signed-off-by: Arnd Bergmann

tee: remove tee_shm_va2pa() and tee_shm_pa2va()

2022-04-26T08:16:48+00:00

We should not need to index into SHMs based on absolute VA/PA. These functions are not used and this kind of usage should not be encouraged anyway. Remove these functions. Signed-off-by: Andrew Davis Reviewed-by: Sumit Garg Signed-off-by: Jens Wiklander

tee: make tee_shm_register_kernel_buf vmalloc supported

2022-04-05T06:04:16+00:00

In some low-memory devices, it's hard to aquire large-orders pages, this patch allowed user using scatter pages to register shm. Signed-off-by: Phil Chang Reviewed-by: Sumit Garg Signed-off-by: Jens Wiklander

tee: refactor TEE_SHM_* flags

2022-02-16T06:49:41+00:00

Removes the redundant TEE_SHM_DMA_BUF, TEE_SHM_EXT_DMA_BUF, TEE_SHM_MAPPED and TEE_SHM_KERNEL_MAPPED flags. TEE_SHM_REGISTER is renamed to TEE_SHM_DYNAMIC in order to better match its usage. Assigns new values to the remaining flags to void gaps. Reviewed-by: Sumit Garg Signed-off-by: Jens Wiklander