[Bug target/61948] [ARM] [4.10 regression] ICE with DImode shift by 1 bit (in copyprop_hardreg_forward_1)

[ramana at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61948

Ramana Radhakrishnan <ramana at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |ice-on-valid-code
                 CC|                            |ramana at gcc dot gnu.org
      Known to work|                            |4.8.3, 4.9.0, 4.9.1
            Version|4.9.0                       |4.10.0
            Summary|[ARM] [4.9 regression] ICE  |[ARM] [4.10 regression] ICE
                   |with DImode shift by 1 bit  |with DImode shift by 1 bit
                   |(in                         |(in
                   |copyprop_hardreg_forward_1) |copyprop_hardreg_forward_1)

--- Comment #1 from Ramana Radhakrishnan <ramana at gcc dot gnu.org> ---
I can't reproduce this on the FSF 4.9 branch but do see it failing on trunk
with Thumb2 and I suspect this to be due to my change to allow any old register
for DImode values in Thumb2. 

Please remember to fill in Keywords, Known to Work, Known to Fail and all
command line options needed to reproduce this issue.


Reproduced on trunk with 
-march=armv7-a -mfloat-abi=hard -mfpu=neon -mthumb