CVE-2026-97920

In the Linux kernel, the following vulnerability has been resolved:

tracing: Keep the entry count when the histogram stats allocation fails

print entries() uses n entries both as the number of sort entries and as its own return value, so the -ENOMEM it stores when the stats allocation fails overwrites the count that the cleanup still needs:

n entries = tracing map sort entries(map, ...);
if (n entries < 0)
	return n entries;
...
		if (!stats) {
			n entries = -ENOMEM;
			goto out;
		}
...

out: tracing map destroy sort entries(sort entries, n entries);

tracing map destroy sort entries() takes an unsigned int and loops up to it, so -ENOMEM arrives as 4294967284. It walks an array of at most map->max elts pointers and calls destroy sort entry(), which dereferences and frees, on whatever lies past the end.

Reading the hist file of a trigger with a .percent value, with that allocation forced to fail:

BUG: KASAN: vmalloc-out-of-bounds in tracing map destroy sort entries+0xa0/0xb0 Read of size 8 at addr ffffc90000045000 by task init/1 tracing map destroy sort entries+0xa0/0xb0 hist show+0x6f7/0x1df0 seq read iter+0x2b8/0x1190 vfs read+0x176/0xa40 The buggy address belongs to a 4-page vmalloc region starting at ffffc90000041000 allocated at tracing map sort entries+0x5c/0xd50

A few pages further the fault is fatal. The registers at the oops confirm the bound: the loop's end pointer less the array start, over the pointer size, is 4294967284.

Return the error in a separate variable and leave n entries holding the count, the way tracing map sort entries() does on its own error path.

The stats block is only entered for a value carrying .percent or .graph, which create val field() has rejected since v6.3, so this cannot be reached in mainline as it stands. It becomes reachable again with "tracing: hist: let values keep the percent and graph modifiers", so it should be applied first.