GNOME Bugzilla – Bug 110226
gnome-vfs-mime-magic.c doesn't compile on HP-UX and Solaris
Last modified: 2004-12-22 21:47:04 UTC
gnome-vfs-mime-magic.c:32:20: wctype.h: No such file or directory gnome-vfs-mime-magic.c: In function `_gnome_vfs_sniff_buffer_looks_like_text': gnome-vfs-mime-magic.c:766: `mbstate_t' undeclared (first use in this function) gnome-vfs-mime-magic.c:766: (Each undeclared identifier is reported only once gnome-vfs-mime-magic.c:766: for each function it appears in.) gnome-vfs-mime-magic.c:766: parse error before "state" gnome-vfs-mime-magic.c:771: `state' undeclared (first use in this function) gmake[7]: *** [gnome-vfs-mime-magic.lo] Error 1 After some googling... it seems that HP-UX 11.00 does not provide that header, perhaps this should be handled by gnome-vfs configure process.
Updating summary. See also http://mail.gnome.org/archives/gnome-vfs-list/2003-April/msg00006.html for a similar problem on Solaris
After some investigation, I found that HP-UX 11i also the correct header. But I dont have any 11i arround here to try compiling. Anyway, there must be a workaround that ./configure could activate. thx
I guess if mbrtowc is not availible we could do an ascii-only text detector.
Created attachment 15650 [details] [review] Patch (not perfect, but should catch compilation problems), needs testers
That is probably good enough. We just need someone to test the patch on a platform with this problem.
Under HP-UX 11.00 Compiles OK!!!! The patch was succesfull. Now I get one error with make ckeck Any Ideas? Im building the rest of gnome right now, perhaps is harmless.... hope so :) -------- CUT ----- Testing get file info... Testing open, close... Testing read, close... Testing cancellation... Testing failed opens... Testing read, cancel, closes... Testing directory loads.** Message: test failed: load directory cancel 1: job never went away delay 10 .** Message: test failed: load directory cancel 2: job never went away delay 0 ** Message: test failed: load directory cancel 1: job never went away delay 1 . Testing directory finds... Testing shutdown... ** (process:17058): WARNING **: ORB: a total of 1 refs to 1 ORB objects were leaked FAIL: test-async-cancel ** (process:17067): WARNING **: ORB: a total of 1 refs to 1 ORB objects were leaked PASS: test-escape PASS: test-uri You should see a warning and 2 critical errors below. (process:17085): libgnomevfs-WARNING **: Attempt to set the thread_count_limit below 2 Trying to start a job with priority = -11; should result in a critical error (process:17085): libgnomevfs-CRITICAL **: file gnome-vfs-async-ops.c: line 643 (gnome_vfs_async_get_file_info): assertion `priority >= GNOME_VFS_PRIORITY_MIN' failed Trying to start a job with priority = 11; should result in a critical error (process:17085): libgnomevfs-CRITICAL **: file gnome-vfs-async-ops.c: line 644 (gnome_vfs_async_get_file_info): assertion `priority <= GNOME_VFS_PRIORITY_MAX' failed No warning or error messages are expected beyond this point warning: failed to load external entity "../test/queue-test-config.xml" Didn't find a valid settings file at ../test/queue-test-config.xml Use the GNOME_VFS_TEST_CONFIG_FILE environment variable to specify a different location. ** (process:17085): WARNING **: ORB: a total of 1 refs to 1 ORB objects were leaked PASS: test-queue Initializing gnome-vfs... Creating async context... Starting open for `./test.input'... Main loop running. Open failed: Invalid URI. Main loop finished. Test async queue efficiency ...Passed All done ** (process:17094): WARNING **: ORB: a total of 1 refs to 1 ORB objects were leaked Test passed PASS: ./auto-test =================== 1 of 5 tests failed =================== gmake[2]: *** [check-TESTS] Error 1 gmake[2]: Leaving directory `/opt/build/garnome-0.22.0/gnome/gnome-vfs/work/main.d/gnome-vfs-2.2.3/test' gmake[1]: *** [check-am] Error 2 gmake[1]: Leaving directory `/opt/build/garnome-0.22.0/gnome/gnome-vfs/work/main.d/gnome-vfs-2.2.3/test' gmake: *** [check-recursive] Error
I committed the fix to both branches Can you file a new bug about the test problem ?
Bug 110907 posted