[Bug target/115148] [12/13/14/15 Regression][SH]: libcanberra fails with 'unaligned opcodes detected in executable segment' since r12-5944-ga7acb6dca941db
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115148 --- Comment #19 from Oleg Endo --- (In reply to John Paul Adrian Glaubitz from comment #18) > This particular bug is resolved when building with an LRA-enabled gcc-15. > > See: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55212 Thanks for checking it. But somehow I have this feeling that this is mere coincidence and the real issue is still lurking somewhere in there.
[Bug target/115148] [12/13/14/15 Regression][SH]: libcanberra fails with 'unaligned opcodes detected in executable segment' since r12-5944-ga7acb6dca941db
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115148 --- Comment #18 from John Paul Adrian Glaubitz --- This particular bug is resolved when building with an LRA-enabled gcc-15. See: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55212
[Bug target/115148] [12/13/14/15 Regression][SH]: libcanberra fails with 'unaligned opcodes detected in executable segment'
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115148 Richard Biener changed: What|Removed |Added Target Milestone|12.4|12.5 --- Comment #17 from Richard Biener --- GCC 12.4 is being released, retargeting bugs to GCC 12.5.
[Bug target/115148] [12/13/14/15 Regression][SH]: libcanberra fails with 'unaligned opcodes detected in executable segment'
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115148 --- Comment #16 from Oleg Endo --- (In reply to John Paul Adrian Glaubitz from comment #15) > Created attachment 58258 [details] > Diff of generated assembly without and with changes from PR99531 > > I have generated a diff that shows the difference in the generated assembly > without and with the patch a7acb6dca941db2b1c135107dac3a34a20650d5c. That's great, thanks a lot! This is the problematic hunks, which causes the wrong code alignment: .LVL108: bt/s.L178 ! mov #-1,r0 !, @@ -1832,36 +1830,39 @@ .byte .L215-.L190 .byte .L181-.L190 .LVL109: - .align 1 -.L192: .LBE111: .LBE110: .loc 1 234 9 .loc 1 234 14 + .align 1 +.L287: + .align 1 +.L288:
[Bug target/115148] [12/13/14/15 Regression][SH]: libcanberra fails with 'unaligned opcodes detected in executable segment'
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115148 --- Comment #15 from John Paul Adrian Glaubitz --- Created attachment 58258 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=58258&action=edit Diff of generated assembly without and with changes from PR99531 I have generated a diff that shows the difference in the generated assembly without and with the patch a7acb6dca941db2b1c135107dac3a34a20650d5c.
[Bug target/115148] [12/13/14/15 Regression][SH]: libcanberra fails with 'unaligned opcodes detected in executable segment'
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115148 Richard Biener changed: What|Removed |Added Keywords||ice-on-valid-code, ra Summary|[SH] [12/13/14 Regression]: |[12/13/14/15 |libcanberra fails with |Regression][SH]: |'unaligned opcodes detected |libcanberra fails with |in executable segment' |'unaligned opcodes detected ||in executable segment' Target Milestone|--- |12.4