Help needed with eglibc 2.13

[Aurelien Jarno <aurelien@aurel32.net>]

Hi alpha porters,

I am trying to get eglibc 2.13 ready to get uploaded into unstable. I
have done the missing porting work on alpha, but I am now stuck with
testsuite regressions, and I would need help to fix them. Here are the
regressions:

- stratcliff.out, Error 1

  This one is caused by the new strncmp() tests which fail on alpha.
  This is not strictly speaking a regression (I mean strncmp() was
  already broken in eglibc 2.11), so I can easily ignore this test if 
  the porters agree.

- tst-cancel24.out, Error 1
- tst-cancelx16.out, Error 1
- tst-cancelx17.out, Error 1
- tst-cancelx18.out, Error 1
- tst-cancelx20.out, Error 1
- tst-cancelx21.out, Error 1
- tst-cancelx4.out, Error 1
- tst-cancelx5.out, Error 1
- tst-cleanupx0.out, Error 1
- tst-cleanupx1.out, Error 1
- tst-cleanupx3.out, Error 1
- tst-cleanupx4.out, Error 1
- tst-mqueue8x.out, Error 1
- tst-oncex3.out, Error 1
- tst-oncex4.out, Error 1

  These tests are built using -fexceptions, which seems to have been
  broken somewhere between eglibc 2.11 and 2.13.

- tst-chk2.out, Error 127
- tst-chk3.out, Error 127
- tst-chk5.out, Error 127
- tst-chk6.out, Error 127
- tst-lfschk2.out, Error 127
- tst-lfschk3.out, Error 127
- tst-lfschk5.out, Error 127
- tst-lfschk6.out, Error 127

  These tests are failing due to the new ____longjmp_chk.S file. In
  eglibc 2.11 given this file was not present, we used to have a file
  containing empty check. In eglibc 2.13 this file is present with alpha
  specific checks, but they seems to be broken as they trigger these
  testsuite errors. We can put back the version used in eglibc 2.11 if
  the porters agree.

All in all if the -fexception issue is solved, we can have a "working"
eglibc on alpha. Help will be greatly appreciated.

Thanks,
Aurelien

-- 
Aurelien Jarno	                        GPG: 1024D/F1BCDB73
aurelien@aurel32.net                 http://www.aurel32.net