diff options
| author | Shivank Garg <shivankg@amd.com> | 2026-08-22 19:22:05 +0000 |
|---|---|---|
| committer | Vinod Koul <vkoul@kernel.org> | 2026-09-09 17:51:44 +0530 |
| commit | e873c74132f0c5f1452816cd9bb26208f0bba1e1 (patch) | |
| tree | cc1ce1567f5b982359331863ebd9da38741f8d2e /rust/ffi.rs | |
| download | linux-stable-e873c74132f0c5f1452816cd9bb26208f0bba1e1.tar.gz linux-stable-e873c74132f0c5f1452816cd9bb26208f0bba1e1.zip | |
dmaengine: fix use-after-free in dma_chan_put() and dma_release_channel()grafted
When dma_device_put() drops the last reference on chan->device->ref,
dma_device_release() runs and may free the dma_device along with its
channels.
dma_chan_put() then still reads chan->device->owner via
dma_chan_to_owner() for the trailing module_put(). KASAN catches it:
slab-use-after-free in dma_chan_put+0x3e6/0x4c0
Read of size 8 by task insmod/6319
Freed by task 6319:
kfree+0x225/0x470
dma_chan_put+0x395/0x4c0
dmaengine_put+0xf8/0x160
Cache the module owner in dma_chan_put() before the put so the trailing
module_put() does not need chan->device.
Fixes: 8ad342a86359 ("dmaengine: Add reference counting to dma_device struct")
Suggested-by: Sashiko <sashiko-bot@kernel.org>
Link: https://sashiko.dev/#/patchset/20260518-dmaengine-kref-fix-v1-1-4d6125048fb7@amd.com
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Shivank Garg <shivankg@amd.com>
Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-3-d4a4ee47d927@amd.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Diffstat (limited to 'rust/ffi.rs')
| -rw-r--r-- | rust/ffi.rs | 50 |
1 files changed, 50 insertions, 0 deletions
diff --git a/rust/ffi.rs b/rust/ffi.rs new file mode 100644 index 000000000..f961e9728 --- /dev/null +++ b/rust/ffi.rs @@ -0,0 +1,50 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Foreign function interface (FFI) types. +//! +//! This crate provides mapping from C primitive types to Rust ones. +//! +//! The Rust [`core`] crate provides [`core::ffi`], which maps integer types to the platform default +//! C ABI. The kernel does not use [`core::ffi`], so it can customise the mapping that deviates from +//! the platform default. + +#![no_std] + +macro_rules! alias { + ($($name:ident = $ty:ty;)*) => {$( + #[allow(non_camel_case_types, missing_docs)] + pub type $name = $ty; + + // Check size compatibility with `core`. + const _: () = assert!( + ::core::mem::size_of::<$name>() == ::core::mem::size_of::<::core::ffi::$name>() + ); + )*} +} + +alias! { + // `core::ffi::c_char` is either `i8` or `u8` depending on architecture. In the kernel, we use + // `-funsigned-char` so it's always mapped to `u8`. + c_char = u8; + + c_schar = i8; + c_uchar = u8; + + c_short = i16; + c_ushort = u16; + + c_int = i32; + c_uint = u32; + + // In the kernel, `intptr_t` is defined to be `long` in all platforms, so we can map the type to + // `isize`. + c_long = isize; + c_ulong = usize; + + c_longlong = i64; + c_ulonglong = u64; +} + +pub use core::ffi::c_void; + +pub use core::ffi::CStr; |
