[Bug target/97251] New: [10/11 Regression] arm: ICEs compiling pure-code/no-literal-pool.c with integer MVE

[acoplan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97251

            Bug ID: 97251
           Summary: [10/11 Regression] arm: ICEs compiling
                    pure-code/no-literal-pool.c with integer MVE
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: acoplan at gcc dot gnu.org
  Target Milestone: ---

Since r10-7193-g63c8f7d6a082b1cd0519fe06d4ed506b04280921, GCC ICEs compiling
gcc/testsuite/gcc.target/arm/pure-code/no-literal-pool.c with integer MVE.

To reproduce:

$ arm-eabi-gcc no-literal-pool.c -c -march=armv8.1-m.main+mve
-mfp16-format=ieee -mpure-code -mfloat-abi=hard
no-literal-pool.c: In function 'testsfp16':
no-literal-pool.c:20:1: error: unrecognizable insn:
   20 | }
      | ^
(insn 33 32 34 2 (set (reg:SF 1 r1)
        (const_double:SF 1.1233999729156494140625e+0 [0x0.8fcb92p+1]))
"no-literal-pool.c":16:6 -1
     (nil))
during RTL pass: vregs
no-literal-pool.c:20:1: internal compiler error: in extract_insn, at
recog.c:2294
0xc6c405 _fatal_insn(char const*, rtx_def const*, char const*, int, char
const*)
        /home/alecop01/toolchain/src/gcc/gcc/rtl-error.c:108
0xc6c424 _fatal_insn_not_found(rtx_def const*, char const*, int, char const*)
        /home/alecop01/toolchain/src/gcc/gcc/rtl-error.c:116
0xc3bf7f extract_insn(rtx_insn*)
        /home/alecop01/toolchain/src/gcc/gcc/recog.c:2294
0x97d49a instantiate_virtual_regs_in_insn
        /home/alecop01/toolchain/src/gcc/gcc/function.c:1607
0x97d49a instantiate_virtual_regs
        /home/alecop01/toolchain/src/gcc/gcc/function.c:1977
0x97d49a execute
        /home/alecop01/toolchain/src/gcc/gcc/function.c:2026
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.

I have a patch (already approved) to fix the issue:
https://gcc.gnu.org/pipermail/gcc-patches/2020-September/555069.html, just
opening the bug to track backporting progress.