On 11/28/13 16:50, Alan Modra wrote:
This is due to that innocuous seeming change of setting
lra_in_progress before calling check_rtl(), in combination with
previous changes Vlad made to the rs6000 backend here:
http://gcc.gnu.org/ml/gcc-patches/2013-10/msg02208.html
In particular the "Call legitimate_constant_pool_address_p in strict
mode for LRA" change, that sets "strict" when lra_in_progress.
Is this still an issue?
That code has gone through a couple revisions. Robert's change was
reverted and Vlad twiddled thigns to use recog_memoized instead of
insn_invalid_p which prevents check_rtl from incorrectly adding CLOBBERs
after the point where an insn's form is supposed to be fixed.
jeff
