On 03/12/24 09:25, Johannes Berg wrote:

On Tue, 2024-12-03 at 23:20 +0800, Haoyu Li wrote:

...

With the new __counted_by annocation in cfg80211_scan_request struct, the "n_channels" struct member must be set before accessing the "channels" array. Failing to do so will trigger a runtime warning when enabling CONFIG_UBSAN_BOUNDS and CONFIG_FORTIFY_SOURCE.

Fixes: e3eac9f32ec0 ("wifi: cfg80211: Annotate struct cfg80211_scan_request with __counted_by")

Signed-off-by: Haoyu Li lihaoyu499@gmail.com

nit: there should be no newline between these

My tolerance for this is going WAY down, it seems it's all just busy- work, and then everyone complains and I need to handle "urgent fixes" because of it etc.

I'm having severe second thoughts about ever having accepted the __counted_by annotations, I think we should just revert it. Experiment failed, we found ... that the code is fine but constantly needs changes to make the checkers happy.

Thanks for taking these changes! - This is improving :) See below.

"Right now, any addition of a counted_by annotation must also include an open-coded assignment of the counter variable after the allocation:

p = alloc(p, array, how_many); p->counter = how_many;

In order to avoid the tedious and error-prone work of manually adding the open-coded counted-by intializations everywhere in the Linux kernel, a new GCC builtin __builtin_counted_by_ref will be very useful to be added to help the adoption of the counted-by attribute.

-- Built-in Function: TYPE __builtin_counted_by_ref (PTR) The built-in function '__builtin_counted_by_ref' checks whether the array object pointed by the pointer PTR has another object associated with it that represents the number of elements in the array object through the 'counted_by' attribute (i.e. the counted-by object). If so, returns a pointer to the corresponding counted-by object. If such counted-by object does not exist, returns a null pointer.

This built-in function is only available in C for now.

The argument PTR must be a pointer to an array. The TYPE of the returned value is a pointer type pointing to the corresponding type of the counted-by object or a void pointer type in case of a null pointer being returned.

With this new builtin, the central allocator could be updated to:

#define MAX(A, B) (A > B) ? (A) : (B) #define alloc(P, FAM, COUNT) ({ \ __auto_type __p = &(P); \ __auto_type __c = (COUNT); \ size_t __size = MAX (sizeof (*(*__p)),\ __builtin_offsetof (__typeof(*(*__p)),FAM) \ + sizeof (*((*__p)->FAM)) * __c); \ if ((*__p = kmalloc(__size))) { \ __auto_type ret = __builtin_counted_by_ref((*__p)->FAM); \ *_Generic(ret, void *: &(size_t){0}, default: ret) = __c; \ } \ })

And then structs can gain the counted_by attribute without needing additional open-coded counter assignments for each struct, and unannotated structs could still use the same allocator."[1]

These changes have been merged already, and will likely be released in coming GCC 15.

For Clang, see [2].

For the kmalloc-family changes, see [3] (a new version of this that includes the __builtin_counted_by_ref() update is coming soon).

-Gustavo

[1] https://gcc.gnu.org/pipermail/gcc-patches/2024-October/665165.html [2] https://github.com/llvm/llvm-project/issues/99774 [3] https://lore.kernel.org/linux-hardening/20240822231324.make.666-kees@kernel....