Alas, the fix for bug 77933 broke __builtin_return_address(0) in both leaf and non-leaf functions on ARM -mthumb, because lr is now more likely to be used as a scratch register to save other high registers, but then its value is used in __builtin_return_address(0) as if it still held the return address. This remains broken in GCC9 (I've just tried r267435). Add "r4" to the clobber list, or compile the original testcase with -fno-omit-frame-pointer, to get: push {r[47], lr} mov lr, r9 mov r[47], r8 push {r[47], lr} [ add r7, sp, #0 ];; -fno-omit-frame-pointer only mov r0, lr Add a call to e.g. bar() between the asm and the return, and you'll also see lr overwritten and then used as if it still held the return address: push {r4, lr} mov lr, r9 mov r4, r8 push {r4, lr} mov r4, lr bl bar movs r0, r4 Before the patch for PR 77933, lr would not be overwritten before use in any of these cases.
Confirmed with -march=armv6-m -mthumb -O2 void baz (void); void * foo (void) { asm volatile("" : : : "r8", "r9", "r4"); return __builtin_return_address (0); } void * bar (void) { asm volatile("" : : : "r8", "r9", "r4"); baz (); return __builtin_return_address (0); }
Could it be that this is a duplicate of bug 88167? I compiled a gcc 7.4.0 patched with the fix for 88167 and now get this result: push {r4, lr} mov r3, r8 mov r4, r9 push {r3, r4} mov r0, lr The patch for bug 88167 seems to be just in trunk for now. As the problem is a regression in all releases till gcc 7 I'd prefer if it could be backported into the corresponding branches.
(In reply to gerd from comment #2) > Could it be that this is a duplicate of bug 88167? > > I compiled a gcc 7.4.0 patched with the fix for 88167 and now get this > result: > > push {r4, lr} > mov r3, r8 > mov r4, r9 > push {r3, r4} > mov r0, lr > > The patch for bug 88167 seems to be just in trunk for now. As the problem is > a regression in all releases till gcc 7 I'd prefer if it could be backported > into the corresponding branches. A regression is not a bug that applies in all previous releases. A regression is where something worked in some previous releases but does not work now.
(In reply to Richard Earnshaw from comment #3) > A regression is not a bug that applies in all previous releases. A > regression is where something worked in some previous releases but does not > work now. using __builtin_return_address(0) as described in the bug report worked in gcc 5.4.0 and before. To be more specific, it worked before the fix for bug 77933 was incorporated. So to me this is a regression.
(In reply to gerd from comment #4) > (In reply to Richard Earnshaw from comment #3) > > A regression is not a bug that applies in all previous releases. A > > regression is where something worked in some previous releases but does not > > work now. > > using __builtin_return_address(0) as described in the bug report worked in > gcc 5.4.0 and before. To be more specific, it worked before the fix for bug > 77933 was incorporated. So to me this is a regression. then please update the known-to-work field accordingly.
(In reply to Richard Earnshaw from comment #5) > then please update the known-to-work field accordingly. It seems I'm missing the rights to do so, could someone with the rights please do that.