Hi,
this is the bootstrap failure of GCC 11 on MinGW64 configured with --enable-
tune=nocona. The bottom line is that SEH does not support CFI for epilogues
but the x86 back-end nevertheless attaches it to instructions, so we have to
filter it out and this is done by detecting the end of the prologue by means
of the NOTE_INSN_PROLOGUE_END note.
But the compiler manages to generate a second epilogue before this note in the
RTL stream and this fools the above logic. The root cause is cross-jumping,
which inserts a jump before the end of the prologue (in fact just before the
note); the rest (CFG cleanup, BB reordering) is downhill from there.
Tested on x86-64/Linux and x86-64/Windows, OK for mainline and 11 branch?
2021-05-05 Eric Botcazou <ebotca...@adacore.com>
PR rtl-optimization/100411
* cfgcleanup.c (try_crossjump_to_edge): Also skip end of prologue
and beginning of function markers.
--
Eric Botcazou
diff --git a/gcc/cfgcleanup.c b/gcc/cfgcleanup.c
index f05cb6136c7..64279cc8c20 100644
--- a/gcc/cfgcleanup.c
+++ b/gcc/cfgcleanup.c
@@ -2148,6 +2148,20 @@ try_crossjump_to_edge (int mode, edge e1, edge e2,
while (DEBUG_INSN_P (newpos1))
newpos1 = NEXT_INSN (newpos1);
+ /* And end of prologue marker. */
+ if (NOTE_P (newpos1) && NOTE_KIND (newpos1) == NOTE_INSN_PROLOGUE_END)
+ newpos1 = NEXT_INSN (newpos1);
+
+ while (DEBUG_INSN_P (newpos1))
+ newpos1 = NEXT_INSN (newpos1);
+
+ /* And also beginning of function marker. */
+ if (NOTE_P (newpos1) && NOTE_KIND (newpos1) == NOTE_INSN_FUNCTION_BEG)
+ newpos1 = NEXT_INSN (newpos1);
+
+ while (DEBUG_INSN_P (newpos1))
+ newpos1 = NEXT_INSN (newpos1);
+
redirect_from = split_block (src1, PREV_INSN (newpos1))->src;
to_remove = single_succ (redirect_from);