[Bug hurd/30146] New: pointless warning about glob64

["bruno at clisp dot org" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=30146

            Bug ID: 30146
           Summary: pointless warning about glob64
           Product: glibc
           Version: 2.34
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: hurd
          Assignee: unassigned at sourceware dot org
          Reporter: bruno at clisp dot org
                CC: samuel.thibault@ens-lyon.org, tschwinge at sourceware dot org
  Target Milestone: ---

Created attachment 14692
  --> https://sourceware.org/bugzilla/attachment.cgi?id=14692&action=edit
unit test

Compiling, on GNU/Hurd, a program that uses the glob() functionality and the
"large file" mode, produces a warning:

warning: glob64 is not implemented and will always fail

But this warning is pointless, as the resulting program works fine.

How to reproduce:
1. Take test-glob.c from the attachment. (It's the 'glob' unit test from
Gnulib.)

2. Compile it:
$ gcc-12 -Wall test-glob.c
/usr/bin/ld: /tmp/ccdgP57r.o: in function `main':
test-glob.c:(.text+0x2d): warning: glob64 is not implemented and will always
fail

3. Look at its symbols:
$ nm a.out
000021bc d _DYNAMIC
000022c4 d _GLOBAL_OFFSET_TABLE_
00000ea8 R _IO_stdin_used
000011b0 r __FRAME_END__
000010ec r __GNU_EH_FRAME_HDR
00002304 D __TMC_END__
00000164 r __abi_tag
00002304 B __bss_start
000022fc D __data_start
000005e0 t __do_global_dtors_aux
000021b8 d __do_global_dtors_aux_fini_array_entry
00002300 D __dso_handle
         U __errno_location@GLIBC_2.2.6
000021b4 d __frame_dummy_init_array_entry
         w __gmon_start__
         U __libc_start_main@GLIBC_2.34
00000550 T __x86.get_pc_thunk.bx
00002304 D _edata
00002308 B _end
00000e90 T _fini
00000ea4 R _fp_hw
0000043c T _init
00000520 T _start
         U abort@GLIBC_2.2.6
         U close@GLIBC_2.2.6
00002304 b completed.0
         U creat64@GLIBC_2.2.6
000022fc W data_start
00000560 t deregister_tm_clones
         U fflush@GLIBC_2.2.6
         U fprintf@GLIBC_2.2.6
00000610 t frame_dummy
         U glob64@GLIBC_2.27
         U globfree64@GLIBC_2.2.6
00000612 T main
000005a0 t register_tm_clones
         U stderr@GLIBC_2.2.6
         U strcmp@GLIBC_2.2.6
         U symlink@GLIBC_2.2.6

You can see that it does use glob64.

4. Run it:
$ ./a.out
$ echo $?
0

It passes the unit test, so the function surely does not "always fail".

-- 
You are receiving this mail because:
You are on the CC list for the bug.