From: Martin Storsjö <martin(a)martin.st>
On ARM and ARM64, each SEH opcode is supposed to map to exactly one
instruction; the operation of writing x29+x30 with a stack offset
with writeback is called ".seh_save_fplr_x".
Using the right opcode would fix inconsistent unwinding if we would
unwind from near the end of the prologue (which is implausible here).
However, clang/LLVM might also get checking for mismatches between
SEH opcodes and the prolog/epilog ranges
(
https://reviews.llvm.org/D131394), mainly to guard against
inconsistencies in the unwind info generated by the compiler, but
that would trigger on this handwritten function too.
---
dlls/wow64/process.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/dlls/wow64/process.c b/dlls/wow64/process.c
index 1d35702bf01..a810d5e5543 100644
--- a/dlls/wow64/process.c
+++ b/dlls/wow64/process.c
@@ -421,8 +421,7 @@ __ASM_GLOBAL_FUNC( raise_exception,
#elif defined(__aarch64__)
__ASM_GLOBAL_FUNC( raise_exception,
"stp x29, x30, [sp, #-32]!\n\t"
- __ASM_SEH(".seh_stackalloc 32\n\t")
- __ASM_SEH(".seh_save_fplr 0\n\t")
+ __ASM_SEH(".seh_save_fplr_x 32\n\t")
__ASM_SEH(".seh_endprologue\n\t")
__ASM_CFI(".cfi_def_cfa x29, 32\n\t")
__ASM_CFI(".cfi_offset x30, -24\n\t")
--
GitLab
https://gitlab.winehq.org/wine/wine/-/merge_requests/998