This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
Hi, The problem here is that we used to allow incrementing casts expression by the extended lvalue rules but we now don't allow this. We now always get an error but the wrong one, as there is no assignment in there. Though this testcase (g++.old-deja/g++.mike/warn1.C) fails on the pointer plus branch because we are trying to build a PLUS_EXPR for pointer types which is wrong. Now this code is dead, I am removing it from the branch and will be submitting the same patch for the mainline but later on after a bootstrap/test there. This brings us down to one testsuite failure in the C++ testsuite which I think was a latent bug in expand. None in the libstdc++ testsuite. There are more failures in libgomp for C++. SCEV still needs to be fixed up too. I have not looked at yet, the Ada, Java, or Fortran front-ends. Committed after a build/test on i686-linux-gnu with no regressions. Thanks, Andrew Pinski ChangeLog: * typeck.c (build_unary_op): Remove code that used to handle non lvalue increments/decrements as we now error out all ways.
Attachment:
fixwarn1.diff.txt
Description: Text document
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |