]> git.ipfire.org Git - thirdparty/glibc.git/blob - BUGS
(gnulib): If have-cc-withh-libunwind is "yes", also mention -lunwind.
[thirdparty/glibc.git] / BUGS
1 List of known bugs (certainly very incomplete)
2 ----------------------------------------------
3
4 Time-stamp: <02/09/30 13:49:48 drepper>
5
6 This following list contains those bugs which I'm aware of. Please
7 make sure that bugs you report are not listed here. If you can fix one
8 of these bugs/limitations I'll certainly be glad to receive a patch.
9
10 Another source of information about bugs is the problem data base of the
11 GNU project. There is an easy to use WWW interface available at
12
13 http://www-gnats.gnu.org:8080/cgi-bin/wwwgnats.pl
14
15 I would appreciate it very much if you could verify the problem was not
16 reported before by looking through the database. To make the information
17 in this database as useful as possible please report bugs always using the
18 `glibcbug' shell script which gets installed with GNU libc. Before reporting
19 a bug please check the FAQ since it discusses also a lot of problematic
20 situations.
21
22
23 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
24
25 Severity: [ *] to [***]
26
27 [ **] Closing shared objects in statically linked binaries most of the
28 times leads to crashes during the dlopen(). Hard to fix.
29
30 [ **] There are problems with signal handling when using LinuxThreads.
31
32 [ **] The RPC code is not 64 bit clean. This is getting slowly fixed
33 but expect incompatible changes on 64 bit platforms like Alpha.
34
35 [ **] If a DSO is using implicitly libpthread and the application itself
36 does not there is a name lookup problem. E.g., the function fork()
37 will be found in the libc.so instead of libpthread since the thread
38 library is behind the libc. To correct this problem it must *not*
39 be relied on the currently still enabled handling of weak symbols
40 in the dynamic linker. Instead explicit tests for the availability
41 of the libpthread version are needed. [PR libc/2325]
42
43 [ *] The precision of the `sinhl' and/or `asinhl' function do not seem
44 to be the best.
45
46 [ *] On Linux, there should be a way to prevent defining the symbol
47 NGROUPS_MAX in the <linux/limits.h> header file. In glibc it
48 is defined in <posix1_lim.h> which must not make the other
49 symbols in <linux/limits.h> available.
50 [PR libc/140]
51
52 [ *] The libm-ieee `gamma' function gives wrong results (at least for
53 -0.5).
54
55 [ *] The libm-ieee `scalb' function gives wrong results for
56 non-integral second parameters.
57
58 [ *] Several (most?) collation specifications are broken. The code which
59 is currently there is in most cases inherited from the originial
60 author (in case there is a LC_COLLATE specification in the locale
61 file) or is defined using the default (if iso14651_t1 is included).
62
63 In any case we are missing information to correct the specification.
64 If you find the specification for your language be faulty please
65 send a report with instruction on what to fix. You don't have to
66 fix the specification yourself.
67
68 The way it finally should look like (if the generic specification
69 is not correct) can be seen in the sv_SE file. Quite a few changes
70 on top of the generic specification can be made without duplication
71 of the whole LC_COLLATE description.
72
73 [ *] Some of the functions which also handled IPv6 are currently broken.
74 IPv6 and IPv4 lookups occasionally happen when not needed. This
75 happens in getaddrinfo() and getnameinfo(). IPv4 handling of
76 these functions is OK though and there are patches available to fix
77 the IPv6 code as well.
78
79 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
80 Ulrich Drepper
81 drepper@redhat.com