[Bug target/117296] [12/13/14/15 Regression] ICE: in expand_gimple_basic_block, at cfgexpand.cc:5896 with -mtune=k6 -mstringop-strategy=libcall -ffloat-store

[pinskia at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117296

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|ICE: in                     |[12/13/14/15 Regression]
                   |expand_gimple_basic_block,  |ICE: in
                   |at cfgexpand.cc:5896 with   |expand_gimple_basic_block,
                   |-mtune=k6                   |at cfgexpand.cc:5896 with
                   |-mstringop-strategy=libcall |-mtune=k6
                   |-ffloat-store               |-mstringop-strategy=libcall
                   |                            |-ffloat-store
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-10-26
      Known to fail|                            |4.9.3, 5.4.0
   Target Milestone|---                         |12.5
             Status|UNCONFIRMED                 |NEW
      Known to work|                            |4.7.1

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
;;   basic block 2, loop depth 0, count 1073741824 (estimated locally), maybe
hot
;;    prev block 3, next block 1, flags: (NEW, RTL, VISITED)
;;    pred:       3 [always]  count:1073741824 (estimated locally) (FALLTHRU)
(insn 14 12 15 (parallel [
            (set (reg/f:SI 7 sp)
                (plus:SI (reg/f:SI 7 sp)
                    (const_int 16 [0x10])))
            (clobber (reg:CC 17 flags))
        ]) "/app/example.cpp":1:24 -1
     (expr_list:REG_ARGS_SIZE (const_int 0 [0])
        (nil)))
(code_label 15 14 0 2 ("l") [0 uses])

Looks like a do_pending_stack_adjust is not in the right place (or missing).
I don't know if this is a target issue or a middle-end issue though.