x86/fpu/xstate: Remove xstate offset check

Traditionally, new xstate components have been assigned sequentially,
aligning feature numbers with their offsets in the XSAVE buffer. However,
this ordering is not architecturally mandated in the non-compacted
format, where a component's offset may not correspond to its feature
number.

The kernel caches CPUID-reported xstate component details, including size
and offset in the non-compacted format. As part of this process, a sanity
check is also conducted to ensure alignment between feature numbers and
offsets.

This check was likely intended as a general guideline rather than a
strict requirement. Upcoming changes will support out-of-order offsets.
Remove the check as becoming obsolete.

Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Link: https://lore.kernel.org/r/20250320234301.8342-2-chang.seok.bae@intel.com
This commit is contained in:
Chang S. Bae 2025-03-20 16:42:52 -07:00 committed by Ingo Molnar
parent 8ffd015db8
commit 031b33ef1a

View File

@ -216,9 +216,6 @@ static bool xfeature_enabled(enum xfeature xfeature)
static void __init setup_xstate_cache(void)
{
u32 eax, ebx, ecx, edx, i;
/* start at the beginning of the "extended state" */
unsigned int last_good_offset = offsetof(struct xregs_state,
extended_state_area);
/*
* The FP xstates and SSE xstates are legacy states. They are always
* in the fixed offsets in the xsave area in either compacted form
@ -246,16 +243,6 @@ static void __init setup_xstate_cache(void)
continue;
xstate_offsets[i] = ebx;
/*
* In our xstate size checks, we assume that the highest-numbered
* xstate feature has the highest offset in the buffer. Ensure
* it does.
*/
WARN_ONCE(last_good_offset > xstate_offsets[i],
"x86/fpu: misordered xstate at %d\n", last_good_offset);
last_good_offset = xstate_offsets[i];
}
}