[Bug target/88892] [8/9 Regression] Double-to-float conversion uses wrong rounding mode when followed by memcpy

[segher at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88892

--- Comment #13 from Segher Boessenkool <segher at gcc dot gnu.org> ---
Author: segher
Date: Fri Jan 18 18:01:56 2019
New Revision: 268083

URL: https://gcc.gnu.org/viewcvs?rev=268083&root=gcc&view=rev
Log:
rs6000: Fix *movsi_from_df (PR88892)

The memory store instructions (stfs[u][x], stxssp[x]) can result in
garbage if the value to be stored isn't already a valid single
precision floating point number.  So we cannot use this here.


        PR target/88892
        * config/rs6000/rs6000.md (*movsi_from_df): Allow only register
        operands.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/rs6000/rs6000.md