On 8/25/25 9:00 AM, Yoshinori Sato wrote:
The problem was caused by an erroneous note about creating a stack frame,
which caused the cur_cfa reg to fail to assert with a value other than
the frame pointer.
This fix will generate notes that correctly update cur_cfa.
gcc/config/rx/
* rx.cc (add_pop_cfi_notes): Release the frame pointer if it is used.
(rx_expand_prologue): Redesigned stack pointer and frame pointer update
process.
So I'd hoped this would fix some of the ~200-ish testsuite failures for
rx that have a similar failure signature but it doesn't seem to have
helped :( On a positive note, it doesn't seem to have caused any new
failures.
As an example of one of the preexisting failures:
FAIL: gcc.dg/torture/pr42363.c -O0 (internal compiler error: in
dwarf2out_frame_debug_adjust_cfa, at dwarf2cfi.cc:1348)
Anyway, it'd be useful to include a testcase with your change. I would
recommend using cvise to simplify the testcase from the PR. Given it's
a fairly straightforward ICE (internal compiler error), I would think
you could likely put the reduced testcase in gcc.dg/torture, possibly
with a dg-additional-options directive to include -fno-omit-frame-pointer.
This will likely help:
https://gcc.gnu.org/wiki/A_guide_to_testcase_reduction
Jeff
