Re: [pdf-devel] FS#108 (and flyspray)

[jemarch]

   >> Can you send a bug report to the 'check' development mailing list,
   > >including details about how to reproduce the problem using libgnupdf?

   Yes, I can. I'll do it asap.

Ok, thanks.

   >   For the sake of completeness, I took these two tests out and ran
   >   everything again. This time everything ran fine, with some of wine's
   >   internal whinings and some test failures, but no page faults. I
   >   attached make's full output and the log from check's run without the
   >   fixtured test cases.
   >
   >> I see that 'pdf_time_from_string_006' is failing.  Are you using a
   > >recent bazaar checkout?

   Mmm. I was using an old bzr checkout. I merged and tried again with
   some news. The whole text suite brings a page fault, but this one's
   not fixed by avoiding the some test cases, have to avoid them all to
   keep it running, it's quite strange, I cannot blame check yet. Just
   commented it out so that I could see what tests were failing. Gosh,
   now I see why we want forked tests!!
   I will pay a look a this later, don't even know how it's working in linux.
   The new list of failures is the following:

   
base/time/pdf-time-from-string.c:474:F:pdf_time_from_string:pdf_time_from_string_006:0:
   Failure 'status != PDF_EBADDATA' occured
   
base/time/pdf-time-to-string.c:210:F:pdf_time_to_string:pdf_time_to_string_002:0:
   Assertion 'strcmp(testStringNA, dateString) == 0' failed

Those failures could be related to the pdf_string global state.

   base/stm/pdf-stm-bseek.c:191:F:pdf_stm_bseek:pdf_stm_bseek_003:0:
   Failure 'ret_char != '3'' occured
   base/stm/pdf-stm-bseek.c:266:F:pdf_stm_bseek:pdf_stm_bseek_004:0:
   Failure 'ret_char != '9'' occured
   base/stm/pdf-stm-write.c:844:F:pdf_stm_write:pdf_stm_write_013:0:
   Failure 'strcmp (data, "GNU") != 0' occured
   base/stm/pdf-stm-write.c:922:F:pdf_stm_write:pdf_stm_write_014:0:
   Failure 'strcmp (data, "GNU") != 0' occured

Quite strange.  The stm module is not depending on any global data, so
the non-forked state of the tests should not affect the result.

-- 
Jose E. Marchesi    address@hidden
GNU Project         http://www.gnu.org