This is the mail archive of the gcc@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, I'm trying to get the LM32 port of GCC 4.8.0 working. Currently the compiler is generating an ICE for a number of newlib files in the same location, when -g and -O1 or -O2 are specified: internal compiler error: in variable_post_merge_new_vals, at var-tracking.c:4430 Which is the following code: gcc_assert (!cselib_lookup (node->loc, GET_MODE (node->loc), 0, VOIDmode)); I'm also seeing this problem on a port to another CPU that isn't in the FSF tree. Any suggestions as to what in the backend might be causing this? i.e. any changes that need to be made from 4.7 (where this problem didn't occur) to 4.8? I've looked through a few other of the simpler ports and I can't really see any changes that have been made that might be related. I don't know much about the var-tracking code though, so not sure where to start looking. I've attached a dump from the last two passes of the compiler. Thanks, Jon
Attachment:
test.tar.gz
Description: Binary data
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |