diff options
| author | Eric Botcazou <ebotcazou@gcc.gnu.org> | 2020-06-15 19:42:11 +0200 |
|---|---|---|
| committer | Eric Botcazou <ebotcazou@gcc.gnu.org> | 2020-06-15 19:54:01 +0200 |
| commit | fb149ebdfee8995ed091f17cd64355ff54e9fb30 (patch) | |
| tree | 4e4998067c3c8211f0e44cc3138318e98ad4bcdb /gcc/gimplify.c | |
| parent | b6ab9ecd550227684643b41e9e33a4d3466724d8 (diff) | |
| download | gcc-fb149ebdfee8995ed091f17cd64355ff54e9fb30.zip gcc-fb149ebdfee8995ed091f17cd64355ff54e9fb30.tar.gz gcc-fb149ebdfee8995ed091f17cd64355ff54e9fb30.tar.bz2 | |
Fix ICE in verify_sra_access_forest
This fixes an issue with reverse storage order in SRA, which is caught
by the built-in verifier in verify_sra_access_forest. The problem is
that propagate_subaccesses_from_rhs changes the type of an access
from aggregate to scalar and, as discussed elsewhere, this must be
done with extra care in the presence of reverse storage order.
gcc/ChangeLog
* tree-sra.c (propagate_subaccesses_from_rhs): When a non-scalar
access on the LHS is replaced with a scalar access, propagate the
TYPE_REVERSE_STORAGE_ORDER flag of the type of the original access.
gcc/testsuite/ChangeLog
* gnat.dg/opt85.ads, gnat.dg/opt85.adb: New test.
Diffstat (limited to 'gcc/gimplify.c')
0 files changed, 0 insertions, 0 deletions