http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58985
Bug ID: 58985 Summary: [4.7 Regression]: gcc.dg/pr57518.c scan-rtl-dump-not ira REG_EQUIV... Product: gcc Version: 4.9.0 Status: UNCONFIRMED Keywords: wrong-code Severity: normal Priority: P3 Component: regression Assignee: unassigned at gcc dot gnu.org Reporter: hp at gcc dot gnu.org CC: vmakarov at gcc dot gnu.org, wmi at google dot com Host: x86_64-unknown-linux-gnu Target: cris-axis-elf This test previously passed, now it fails. Revision r204212 (removing the regmove pass) caused this test to fail for cris-elf. This PR tracks this regression. The situation was problematic because at the time, the tree was broken due to revision r204205, reverted in r204254. In this revision range the regression was tracked with the vital part of r204254 applied; the test still fails at r204346. In gcc.log: PASS: gcc.dg/pr57518.c (test for excess errors) PASS: gcc.dg/pr57518.c scan-rtl-dump-not ira "REG_EQUIV[^\n]*mem[^\n]*"ip"" I have some doubts that the test-case is actually valid as-is for all ports, see <http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57518#c8>. Author of revision 204212 and author of PR57518 CC:ed.