On 8/1/2010 10:42 AM, Paolo Bonzini wrote: > This fixes 37 and 38 (and likely a bunch of other failures).
Yep. > The 37/38 failure is really a logic bug, the other two are cut-and-paste > problems. It seems like I didn't review the patch even. (I did; what > I didn't do was testing it beyond sysroot.at :). > > I'll squash it in 3/9. > Briefly tested in conjunction with this: http://lists.gnu.org/archive/html/libtool-patches/2010-07/msg00127.html That is, I only ran part of the test suite. I'll do a full run on cygwin-native using the v4 series (withOUT the extra patch above) later today, and report those results. (Thanks to Ralf for running the full suite on linux, I don't need to do that. And I really only want to run the full suite on MinGW/MSYS once...'cause it's still a six hour attendance-required ordeal to manually dismiss popups. So, I'll defer THAT torture until things settle down after Ralf's review comments are incorporated.) Current preliminary results (v3 + Paolo's fixup in this thread + my/Paolo's fixup referenced above): cygwin-native: Tested old:tagdemo-shared+old:tagdemo-make; new:37,38,41,101,117-119. Results: all passed (or skipped). linux-native: Tested old:tagdemo-shared+old:tagdemo-make; new:37,38,41,101,117-119. Results: all passed (or skipped). mingw native (MSYS hosted): Tested: old:tagdemo-shared+old:tagdemo-make; new:37,38,41,101,117-119. Results: all passed (or skipped) cygwin->mingw cross: Tested: new:117-119 Results: all passed linux->mingw cross Tested: new:117-119 Results: all passed -- Chuck
