diff options
| author | Richard Sandiford <richard.sandiford@linaro.org> | 2018-03-01 08:22:06 +0000 |
|---|---|---|
| committer | Richard Sandiford <rsandifo@gcc.gnu.org> | 2018-03-01 08:22:06 +0000 |
| commit | e6eacdc9451bbee54c80d6b85d22ba390fd2c7c4 (patch) | |
| tree | 351106ed56d48df57047be4534967c2e4d1300b3 /gcc/fortran/trans-array.c | |
| parent | 54178a01107b911813609693dd8d91968ac06819 (diff) | |
| download | gcc-e6eacdc9451bbee54c80d6b85d22ba390fd2c7c4.zip gcc-e6eacdc9451bbee54c80d6b85d22ba390fd2c7c4.tar.gz gcc-e6eacdc9451bbee54c80d6b85d22ba390fd2c7c4.tar.bz2 | |
Tighten use of HARD_FRAME_POINTER_REGNUM in alias.c (PR 84538)
RTL code needs to be consistent about whether it uses the stack
pointer, the frame pointer or the argument pointer to access a
given part of the frame. alias.c used this to divide accesses
into three independent areas.
The problem in the PR is that we did this for HARD_FRAME_POINTER_REGNUM
even when the register wasn't being used as a frame pointer. We can't
do that because the frame pointer is then just any old allocatable
register and could certainly point to info accessed through the
argument pointer or stack pointer.
2018-03-01 Richard Sandiford <richard.sandiford@linaro.org>
gcc/
PR rtl-optimization/84538
* alias.c (init_alias_target): Add commentary.
(init_alias_analysis): Only give HARD_FRAME_POINTER_REGNUM
a unique base value if the frame pointer is not eliminated
to the stack pointer.
gcc/testsuite/
PR rtl-optimization/84538
* gcc.dg/torture/pr84538.c: New test.
From-SVN: r258094
Diffstat (limited to 'gcc/fortran/trans-array.c')
0 files changed, 0 insertions, 0 deletions