[Bug debug/107183] New: -fcompare-debug failure (length) with -fsanitize=float-cast-overflow

[Bug debug/107183] New: -fcompare-debug failure (length) with -fsanitize=float-cast-overflow

[zsojka at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

            Bug ID: 107183
           Summary: -fcompare-debug failure (length) with
                    -fsanitize=float-cast-overflow
           Product: gcc
           Version: 13.0
            Status: UNCONFIRMED
          Keywords: compare-debug-failure
          Severity: normal
          Priority: P3
         Component: debug
          Assignee: unassigned at gcc dot gnu.org
          Reporter: zsojka at seznam dot cz
                CC: aoliva at gcc dot gnu.org
  Target Milestone: ---
              Host: x86_64-pc-linux-gnu
            Target: x86_64-pc-linux-gnu

Created attachment 53678
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53678&action=edit
reduced testcase

Compiler output:
$ x86_64-pc-linux-gnu-gcc -O -fsanitize=float-cast-overflow -fcompare-debug
testcase.c
x86_64-pc-linux-gnu-gcc: error: testcase.c: '-fcompare-debug' failure (length)
$ diff -u *gkd
--- a-testcase.c.gkd    2022-10-07 20:13:29.246165247 +0200
+++ a-testcase.gk.c.gkd 2022-10-07 20:13:29.299498288 +0200
@@ -79,12 +79,19 @@
         (const_double:SF 0.0 [0x0.0p+0])) "testcase.c":6:18 discrim 2#
{*movsf_internal}
      (expr_list:REG_EQUIV (const_double:SF 0.0 [0x0.0p+0])
         (nil)))
+(insn # 0 0 4 (parallel [
+            (set (reg:XF 9 st(1))
+                (reg:XF 8 st))
+            (set (reg:XF 8 st)
+                (reg:XF 9 st(1)))
+        ]) "testcase.c":6:18 discrim 2# {*swapxf}
+     (nil))
 (insn # 0 0 4 (set (reg:CCFP 17 flags)
         (unspec:CCFP [
                 (compare:CCFP (reg:XF 8 st)
                     (reg:XF 9 st(1)))
             ] UNSPEC_NOTRAP)) "testcase.c":6:18 discrim 2# {*cmpiuxf_i387}
-     (expr_list:REG_DEAD (reg:XF 8 st)
+     (expr_list:REG_DEAD (reg:XF 8 st [107])
         (nil)))
 (insn # 0 0 4 (set (reg:XF 8 st)
         (reg:XF 8 st)) "testcase.c":6:18 discrim 2# {*movxf_internal}


$ x86_64-pc-linux-gnu-gcc -v
Using built-in specs.
COLLECT_GCC=/repo/gcc-trunk/binary-latest-amd64/bin/x86_64-pc-linux-gnu-gcc
COLLECT_LTO_WRAPPER=/repo/gcc-trunk/binary-trunk-r13-3161-20221007142737-g89228e3985c-checking-yes-rtl-df-extra-nobootstrap-amd64/bin/../libexec/gcc/x86_64-pc-linux-gnu/13.0.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /repo/gcc-trunk//configure --enable-languages=c,c++
--enable-valgrind-annotations --disable-nls --enable-checking=yes,rtl,df,extra
--disable-bootstrap --with-cloog --with-ppl --with-isl
--build=x86_64-pc-linux-gnu --host=x86_64-pc-linux-gnu
--target=x86_64-pc-linux-gnu --with-ld=/usr/bin/x86_64-pc-linux-gnu-ld
--with-as=/usr/bin/x86_64-pc-linux-gnu-as --disable-libstdcxx-pch
--prefix=/repo/gcc-trunk//binary-trunk-r13-3161-20221007142737-g89228e3985c-checking-yes-rtl-df-extra-nobootstrap-amd64
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 13.0.0 20221007 (experimental) (GCC)

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|debug                       |target
            Summary|-fcompare-debug failure     |[10/11/12/13 Regression]
                   |(length) with               |-fcompare-debug failure
                   |-fsanitize=float-cast-overf |(length) with
                   |low                         |-fsanitize=float-cast-overf
                   |                            |low
   Target Milestone|---                         |10.5

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I am 99% sure it is a bug in reg-stack.cc which makes this a target issue as
x86_64 is one of two targets which use reg-stack for their FP (m68k being the
other).

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2022-10-10
             Status|UNCONFIRMED                 |NEW
            Summary|[10/11/12/13 Regression]    |[10/11/12/13 Regression]
                   |-fcompare-debug failure     |-fcompare-debug failure
                   |(length) with               |(length) with
                   |-fsanitize=float-cast-overf |-fsanitize=float-cast-overf
                   |low                         |low since
                   |                            |r7-5708-gcfd719e7769fd43f
                 CC|                            |jakub at gcc dot gnu.org,
                   |                            |marxin at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Likely started with r7-5708-gcfd719e7769fd43f.

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Martin Liška from comment #2)
> Likely started with r7-5708-gcfd719e7769fd43f.

Exposed by ....

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P2

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Seems the bug is in swap_rtx_condition.
It is called on:
(insn 10023 10021 10024 4 (set (reg:CCFP 17 flags)
        (unspec:CCFP [
                (compare:CCFP (reg:XF 9 st(1) [orig:84 _3 ] [84])
                    (reg:XF 8 st [107]))
            ] UNSPEC_NOTRAP)) "pr107183.c":6:18 discrim 1 45 {*cmpiuxf_i387}
     (expr_list:REG_DEAD (reg:XF 9 st(1) [orig:84 _3 ] [84])
        (expr_list:REG_DEAD (reg:XF 8 st [107])
            (nil))))
on the first next_flags_user call in the function it sets insn to:
(insn 10110 10024 10111 4 (set (reg:QI 3 bx [105])
        (unordered:QI (reg:CCFP 17 flags)
            (const_int 0 [0]))) "pr107183.c":6:18 discrim 1 949 {*setcc_qi}
     (nil))
(still the same between -g and -g0, but already that one could differ),
but then next_flags_user (the 3rd one in the function) finds
(debug_insn 4 10029 10030 4 (var_location:SI D#1 (if_then_else:SI (uneq
(reg:CCFP 17 flags)
            (const_int 0 [0]))
        (reg:SI 3 bx [105])
        (reg:SI 0 ax [111]))) -1
     (nil))
with -g and
(insn 10030 10029 10032 4 (set (reg:SI 3 bx [orig:90 iftmp.0_9 ] [90])
        (if_then_else:SI (uneq (reg:CCFP 17 flags)
                (const_int 0 [0]))
            (reg:SI 3 bx [105])
            (reg:SI 0 ax [111]))) "pr107183.c":6:18 discrim 1 1269
{*movsicc_noc}
     (expr_list:REG_DEAD (reg:CCFP 17 flags)
        (expr_list:REG_DEAD (reg:SI 0 ax [111])
            (expr_list:REG_EQUAL (if_then_else:SI (uneq (reg:CCFP 17 flags)
                        (const_int 0 [0]))
                    (reg:SI 3 bx [105])
                    (const_int 1 [0x1]))
                (nil)))))
with -g0.

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |jakub at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 53926
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53926&action=edit
gcc13-pr107183.patch

Untested fix.

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:6b5c98c1c0003bd470a4428bede6c862637a94b8

commit r13-4175-g6b5c98c1c0003bd470a4428bede6c862637a94b8
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Sun Nov 20 17:42:42 2022 +0100

    reg-stack: Fix a -fcompare-debug bug in reg-stack [PR107183]

    As the following testcase shows, the swap_rtx_condition function
    in reg-stack can result in different code generation between -g and -g0.
    The function is doing the changes as it goes, so does analysis and
    changes together, which makes it harder to deal with DEBUG_INSNs,
    where normally analysis phase ignores them and the later phase
    doesn't.
    swap_rtx_condition walks instructions two different ways, one is
    using next_flags_user function which stops on non-call instructions
    that mention the flags register, and the other is a loop on fnstsw
    where it stops on instructions mentioning it and tries to find
    sahf instruction that uses it (in both cases calls stop it and so
    does end of basic block).
    Now both of these currently stop on DEBUG_INSNs that mention
    the flags register resp. the fnstsw result register.
    On success the function recurses on next flags user instruction
    if still live and if the recursion failed, reverts the changes
    it did too and fails.
    If it were just for the next_flags_user case, the fix could be
    just not doing
          INSN_CODE (insn) = -1;
          if (recog_memoized (insn) == -1)
            fail = 1;
    on DEBUG_INSNs (assuming all changes to those are fine),
    swap_rtx_condition_1 just changes one comparison to a different
    one.  But due to the possibility of fnstsw result being used
    in theory before sahf in some DEBUG_INSNs, this patch takes
    a different approach.  swap_rtx_condition has now a new argument
    and two modes.  The first mode is when debug_seen is >= 0, in this
    case both next_flags_user and the loop for fnstsw -> sahf will
    ignore but note DEBUG_INSNs (that mention flags register or fnstsw
    result).  If no such DEBUG_INSN is found during the whole call
    including recursive invocations (so e.g. for -g0 but probably most
    often for -g as well), it behaves as before, if it returns true
    all the changes are done and nothing further needs to be done later.
    If any DEBUG_INSNs are seen along the way, even when returning success
    all the changes are reverted, so it just reports that the function
    would be successful if DEBUG_INSNs were ignored.
    In this case, compare_for_stack_reg needs to call it again in
    debug_seen = -1 mode, which tells the function to update everything
    including DEBUG_INSNs.  For the fnstsw -> sahf case which I hope
    will be very rare I just reset the DEBUG_INSNs, I don't really
    know how to express it easily otherwise.  For the rest
    swap_rtx_condition_1 is done even on the DEBUG_INSNs.

    2022-11-20  Jakub Jelinek  <jakub@redhat.com>

            PR target/107183
            * reg-stack.cc (next_flags_user): Add DEBUG_SEEN argument.
            If >= 0 and a DEBUG_INSN would be otherwise returned, set
            DEBUG_SEEN to 1 and ignore it.
            (swap_rtx_condition): Add DEBUG_SEEN argument.  In >= 0
            mode only set DEBUG_SEEN to 1 if problematic DEBUG_ISNSs
            were seen and revert all changes on success in that case.
            Don't try to recog_memoized DEBUG_INSNs.
            (compare_for_stack_reg): Adjust swap_rtx_condition caller.
            If it returns true and debug_seen is 1, call swap_rtx_condition
            again with debug_seen -1.

            * gcc.dg/ubsan/pr107183.c: New test.

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

--- Comment #7 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-12 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:196cb25ddbc6c4fae5e7b1c0a09f32f59bdabaa8

commit r12-8922-g196cb25ddbc6c4fae5e7b1c0a09f32f59bdabaa8
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Sun Nov 20 17:42:42 2022 +0100

    reg-stack: Fix a -fcompare-debug bug in reg-stack [PR107183]

    As the following testcase shows, the swap_rtx_condition function
    in reg-stack can result in different code generation between -g and -g0.
    The function is doing the changes as it goes, so does analysis and
    changes together, which makes it harder to deal with DEBUG_INSNs,
    where normally analysis phase ignores them and the later phase
    doesn't.
    swap_rtx_condition walks instructions two different ways, one is
    using next_flags_user function which stops on non-call instructions
    that mention the flags register, and the other is a loop on fnstsw
    where it stops on instructions mentioning it and tries to find
    sahf instruction that uses it (in both cases calls stop it and so
    does end of basic block).
    Now both of these currently stop on DEBUG_INSNs that mention
    the flags register resp. the fnstsw result register.
    On success the function recurses on next flags user instruction
    if still live and if the recursion failed, reverts the changes
    it did too and fails.
    If it were just for the next_flags_user case, the fix could be
    just not doing
          INSN_CODE (insn) = -1;
          if (recog_memoized (insn) == -1)
            fail = 1;
    on DEBUG_INSNs (assuming all changes to those are fine),
    swap_rtx_condition_1 just changes one comparison to a different
    one.  But due to the possibility of fnstsw result being used
    in theory before sahf in some DEBUG_INSNs, this patch takes
    a different approach.  swap_rtx_condition has now a new argument
    and two modes.  The first mode is when debug_seen is >= 0, in this
    case both next_flags_user and the loop for fnstsw -> sahf will
    ignore but note DEBUG_INSNs (that mention flags register or fnstsw
    result).  If no such DEBUG_INSN is found during the whole call
    including recursive invocations (so e.g. for -g0 but probably most
    often for -g as well), it behaves as before, if it returns true
    all the changes are done and nothing further needs to be done later.
    If any DEBUG_INSNs are seen along the way, even when returning success
    all the changes are reverted, so it just reports that the function
    would be successful if DEBUG_INSNs were ignored.
    In this case, compare_for_stack_reg needs to call it again in
    debug_seen = -1 mode, which tells the function to update everything
    including DEBUG_INSNs.  For the fnstsw -> sahf case which I hope
    will be very rare I just reset the DEBUG_INSNs, I don't really
    know how to express it easily otherwise.  For the rest
    swap_rtx_condition_1 is done even on the DEBUG_INSNs.

    2022-11-20  Jakub Jelinek  <jakub@redhat.com>

            PR target/107183
            * reg-stack.cc (next_flags_user): Add DEBUG_SEEN argument.
            If >= 0 and a DEBUG_INSN would be otherwise returned, set
            DEBUG_SEEN to 1 and ignore it.
            (swap_rtx_condition): Add DEBUG_SEEN argument.  In >= 0
            mode only set DEBUG_SEEN to 1 if problematic DEBUG_ISNSs
            were seen and revert all changes on success in that case.
            Don't try to recog_memoized DEBUG_INSNs.
            (compare_for_stack_reg): Adjust swap_rtx_condition caller.
            If it returns true and debug_seen is 1, call swap_rtx_condition
            again with debug_seen -1.

            * gcc.dg/ubsan/pr107183.c: New test.

    (cherry picked from commit 6b5c98c1c0003bd470a4428bede6c862637a94b8)

[Bug target/107183] [10/11/12/13 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-11 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:d92cbc49eac263e32ff5eaabffd3efe2324502d2

commit r11-10386-gd92cbc49eac263e32ff5eaabffd3efe2324502d2
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Sun Nov 20 17:42:42 2022 +0100

    reg-stack: Fix a -fcompare-debug bug in reg-stack [PR107183]

    As the following testcase shows, the swap_rtx_condition function
    in reg-stack can result in different code generation between -g and -g0.
    The function is doing the changes as it goes, so does analysis and
    changes together, which makes it harder to deal with DEBUG_INSNs,
    where normally analysis phase ignores them and the later phase
    doesn't.
    swap_rtx_condition walks instructions two different ways, one is
    using next_flags_user function which stops on non-call instructions
    that mention the flags register, and the other is a loop on fnstsw
    where it stops on instructions mentioning it and tries to find
    sahf instruction that uses it (in both cases calls stop it and so
    does end of basic block).
    Now both of these currently stop on DEBUG_INSNs that mention
    the flags register resp. the fnstsw result register.
    On success the function recurses on next flags user instruction
    if still live and if the recursion failed, reverts the changes
    it did too and fails.
    If it were just for the next_flags_user case, the fix could be
    just not doing
          INSN_CODE (insn) = -1;
          if (recog_memoized (insn) == -1)
            fail = 1;
    on DEBUG_INSNs (assuming all changes to those are fine),
    swap_rtx_condition_1 just changes one comparison to a different
    one.  But due to the possibility of fnstsw result being used
    in theory before sahf in some DEBUG_INSNs, this patch takes
    a different approach.  swap_rtx_condition has now a new argument
    and two modes.  The first mode is when debug_seen is >= 0, in this
    case both next_flags_user and the loop for fnstsw -> sahf will
    ignore but note DEBUG_INSNs (that mention flags register or fnstsw
    result).  If no such DEBUG_INSN is found during the whole call
    including recursive invocations (so e.g. for -g0 but probably most
    often for -g as well), it behaves as before, if it returns true
    all the changes are done and nothing further needs to be done later.
    If any DEBUG_INSNs are seen along the way, even when returning success
    all the changes are reverted, so it just reports that the function
    would be successful if DEBUG_INSNs were ignored.
    In this case, compare_for_stack_reg needs to call it again in
    debug_seen = -1 mode, which tells the function to update everything
    including DEBUG_INSNs.  For the fnstsw -> sahf case which I hope
    will be very rare I just reset the DEBUG_INSNs, I don't really
    know how to express it easily otherwise.  For the rest
    swap_rtx_condition_1 is done even on the DEBUG_INSNs.

    2022-11-20  Jakub Jelinek  <jakub@redhat.com>

            PR target/107183
            * reg-stack.c (next_flags_user): Add DEBUG_SEEN argument.
            If >= 0 and a DEBUG_INSN would be otherwise returned, set
            DEBUG_SEEN to 1 and ignore it.
            (swap_rtx_condition): Add DEBUG_SEEN argument.  In >= 0
            mode only set DEBUG_SEEN to 1 if problematic DEBUG_ISNSs
            were seen and revert all changes on success in that case.
            Don't try to recog_memoized DEBUG_INSNs.
            (compare_for_stack_reg): Adjust swap_rtx_condition caller.
            If it returns true and debug_seen is 1, call swap_rtx_condition
            again with debug_seen -1.

            * gcc.dg/ubsan/pr107183.c: New test.

    (cherry picked from commit 6b5c98c1c0003bd470a4428bede6c862637a94b8)

[Bug target/107183] [10 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11/12/13 Regression]    |[10 Regression]
                   |-fcompare-debug failure     |-fcompare-debug failure
                   |(length) with               |(length) with
                   |-fsanitize=float-cast-overf |-fsanitize=float-cast-overf
                   |low since                   |low since
                   |r7-5708-gcfd719e7769fd43f   |r7-5708-gcfd719e7769fd43f

--- Comment #9 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed for 11.4+, 12.3+ and trunk so far.

[Bug target/107183] [10 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

--- Comment #10 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:2c83256249624cdf6feae6814cc31ef2cd94c185

commit r10-11342-g2c83256249624cdf6feae6814cc31ef2cd94c185
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Sun Nov 20 17:42:42 2022 +0100

    reg-stack: Fix a -fcompare-debug bug in reg-stack [PR107183]

    As the following testcase shows, the swap_rtx_condition function
    in reg-stack can result in different code generation between -g and -g0.
    The function is doing the changes as it goes, so does analysis and
    changes together, which makes it harder to deal with DEBUG_INSNs,
    where normally analysis phase ignores them and the later phase
    doesn't.
    swap_rtx_condition walks instructions two different ways, one is
    using next_flags_user function which stops on non-call instructions
    that mention the flags register, and the other is a loop on fnstsw
    where it stops on instructions mentioning it and tries to find
    sahf instruction that uses it (in both cases calls stop it and so
    does end of basic block).
    Now both of these currently stop on DEBUG_INSNs that mention
    the flags register resp. the fnstsw result register.
    On success the function recurses on next flags user instruction
    if still live and if the recursion failed, reverts the changes
    it did too and fails.
    If it were just for the next_flags_user case, the fix could be
    just not doing
          INSN_CODE (insn) = -1;
          if (recog_memoized (insn) == -1)
            fail = 1;
    on DEBUG_INSNs (assuming all changes to those are fine),
    swap_rtx_condition_1 just changes one comparison to a different
    one.  But due to the possibility of fnstsw result being used
    in theory before sahf in some DEBUG_INSNs, this patch takes
    a different approach.  swap_rtx_condition has now a new argument
    and two modes.  The first mode is when debug_seen is >= 0, in this
    case both next_flags_user and the loop for fnstsw -> sahf will
    ignore but note DEBUG_INSNs (that mention flags register or fnstsw
    result).  If no such DEBUG_INSN is found during the whole call
    including recursive invocations (so e.g. for -g0 but probably most
    often for -g as well), it behaves as before, if it returns true
    all the changes are done and nothing further needs to be done later.
    If any DEBUG_INSNs are seen along the way, even when returning success
    all the changes are reverted, so it just reports that the function
    would be successful if DEBUG_INSNs were ignored.
    In this case, compare_for_stack_reg needs to call it again in
    debug_seen = -1 mode, which tells the function to update everything
    including DEBUG_INSNs.  For the fnstsw -> sahf case which I hope
    will be very rare I just reset the DEBUG_INSNs, I don't really
    know how to express it easily otherwise.  For the rest
    swap_rtx_condition_1 is done even on the DEBUG_INSNs.

    2022-11-20  Jakub Jelinek  <jakub@redhat.com>

            PR target/107183
            * reg-stack.c (next_flags_user): Add DEBUG_SEEN argument.
            If >= 0 and a DEBUG_INSN would be otherwise returned, set
            DEBUG_SEEN to 1 and ignore it.
            (swap_rtx_condition): Add DEBUG_SEEN argument.  In >= 0
            mode only set DEBUG_SEEN to 1 if problematic DEBUG_ISNSs
            were seen and revert all changes on success in that case.
            Don't try to recog_memoized DEBUG_INSNs.
            (compare_for_stack_reg): Adjust swap_rtx_condition caller.
            If it returns true and debug_seen is 1, call swap_rtx_condition
            again with debug_seen -1.

            * gcc.dg/ubsan/pr107183.c: New test.

    (cherry picked from commit 6b5c98c1c0003bd470a4428bede6c862637a94b8)

[Bug target/107183] [10 Regression] -fcompare-debug failure (length) with -fsanitize=float-cast-overflow since r7-5708-gcfd719e7769fd43f

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107183

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #11 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed for 10.5 too.