LTO and libelf (and FreeBSD)

Gerald Pfeifer gerald@pfeifer.com
Sun May 23 15:16:00 GMT 2010


On Sat, 22 May 2010, Gerald Pfeifer wrote:
> I'll be submitting result for that around noon your time tomorrow-
> Right now I am testing vanilla GCC and patched FreeBSD libelf, my
> tester is just rather slow.

Like Kai's patch to FreeBSD's libelf 
  http://gcc.gnu.org/ml/gcc-testresults/2010-05/msg02150.html
your improves results from 
http://gcc.gnu.org/ml/gcc-testresults/2010-05/msg02081.html
to http://gcc.gnu.org/ml/gcc-testresults/2010-05/msg02178.html .

=== g++ Summary ===		vanilla	patched	Richard's
                   			libelf	GCC patch

# of expected passes		22876	23118	23118
# of unexpected failures	425	142	142
# of expected failures		151	151	151
# of unresolved testcases	124	66	66
# of unsupported tests		150	150	150

=== gcc Summary ===

# of expected passes		64021	67099	67099
# of unexpected failures	3806	1667	1667
# of unexpected successes	1	1	1
# of expected failures		171	171	171
# of unresolved testcases	2580	1292	1292
# of unsupported tests		1284	1284	1284

=== gfortran Summary === (one test added for the last run)

# of expected passes		34094	34122	34123
# of unexpected failures	48	27	27
# of expected failures		22	22	22
# of unresolved testcases	37	16	16
# of unsupported tests		266	266	266


In fact (not surprisingly, looking at the code) both patches have the same 
effect and both the one to FreeBSD libelf and GCC should be applied.


What's missing then are the -fwhopr failures, which appear to be the
same in both cases:

  % cat x.c
  int main() { }
  % gccvs -flto x.c
  % gccvs -fwhopr x.c
  lto1: fatal error: elf_update() failed: Layout constraint violation
  compilation terminated.
  lto-wrapper: gccvs returned 1 exit status
  collect2: lto-wrapper returned 1 exit status

Gerald



More information about the Gcc mailing list