]>
Commit | Line | Data |
---|---|---|
61952351 | 1 | Frequently Asked Questions about the GNU C Library |
f8cac037 | 2 | |
f12944ec UD |
3 | This document tries to answer questions a user might have when installing |
4 | and using glibc. Please make sure you read this before sending questions or | |
5 | bug reports to the maintainers. | |
f8cac037 | 6 | |
f12944ec | 7 | The GNU C library is very complex. The installation process has not been |
fdacb17d | 8 | completely automated; there are too many variables. You can do substantial |
f12944ec UD |
9 | damage to your system by installing the library incorrectly. Make sure you |
10 | understand what you are undertaking before you begin. | |
f8cac037 | 11 | |
41f27456 RM |
12 | If you have any questions you think should be answered in this document, |
13 | please let me know. | |
f8cac037 RM |
14 | |
15 | --drepper@cygnus.com | |
16 | \f | |
61952351 UD |
17 | ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ |
18 | ||
19 | 1. Compiling glibc | |
20 | ||
21 | 1.1. What systems does the GNU C Library run on? | |
22 | 1.2. What compiler do I need to build GNU libc? | |
23 | 1.3. When I try to compile glibc I get only error messages. | |
24 | What's wrong? | |
5edb9387 | 25 | 1.4. Do I need a special linker or assembler? |
8619129f | 26 | 1.5. Which compiler should I use for powerpc? |
348ed515 UD |
27 | 1.6. Which tools should I use for ARM? |
28 | 1.7. Do I need some more things to compile the GNU C Library? | |
29 | 1.8. What version of the Linux kernel headers should be used? | |
30 | 1.9. The compiler hangs while building iconvdata modules. What's | |
f12944ec | 31 | wrong? |
348ed515 | 32 | 1.10. When I run `nm -u libc.so' on the produced library I still |
61952351 | 33 | find unresolved symbols. Can this be ok? |
348ed515 UD |
34 | 1.11. What are these `add-ons'? |
35 | 1.12. My XXX kernel emulates a floating-point coprocessor for me. | |
61952351 | 36 | Should I enable --with-fp? |
348ed515 | 37 | 1.13. When compiling GNU libc I get lots of errors saying functions |
61952351 | 38 | in glibc are duplicated in libgcc. |
348ed515 | 39 | 1.14. Why do I get messages about missing thread functions when I use |
a35cb74d | 40 | librt? I don't even use threads. |
348ed515 | 41 | 1.15. What's the problem with configure --enable-omitfp? |
b1418d8f | 42 | 1.16. I get failures during `make check'. What should I do? |
348ed515 | 43 | 1.17. What is symbol versioning good for? Do I need it? |
61952351 UD |
44 | |
45 | 2. Installation and configuration issues | |
46 | ||
47 | 2.1. Can I replace the libc on my Linux system with GNU libc? | |
48 | 2.2. How do I configure GNU libc so that the essential libraries | |
49 | like libc.so go into /lib and the other into /usr/lib? | |
50 | 2.3. How should I avoid damaging my system when I install GNU libc? | |
51 | 2.4. Do I need to use GNU CC to compile programs that will use the | |
52 | GNU C Library? | |
53 | 2.5. When linking with the new libc I get unresolved symbols | |
54 | `crypt' and `setkey'. Why aren't these functions in the | |
55 | libc anymore? | |
56 | 2.6. When I use GNU libc on my Linux system by linking against | |
57 | the libc.so which comes with glibc all I get is a core dump. | |
58 | 2.7. Looking through the shared libc file I haven't found the | |
59 | functions `stat', `lstat', `fstat', and `mknod' and while | |
60 | linking on my Linux system I get error messages. How is | |
61 | this supposed to work? | |
5edb9387 UD |
62 | 2.8. When I run an executable on one system which I compiled on |
63 | another, I get dynamic linker errors. Both systems have the same | |
64 | version of glibc installed. What's wrong? | |
65 | 2.9. How can I compile gcc 2.7.2.1 from the gcc source code using | |
61952351 | 66 | glibc 2.x? |
5edb9387 | 67 | 2.10. The `gencat' utility cannot process the catalog sources which |
61952351 | 68 | were used on my Linux libc5 based system. Why? |
5edb9387 | 69 | 2.11. Programs using libc have their messages translated, but other |
a35cb74d | 70 | behavior is not localized (e.g. collating order); why? |
5edb9387 | 71 | 2.12. I have set up /etc/nis.conf, and the Linux libc 5 with NYS |
61952351 | 72 | works great. But the glibc NIS+ doesn't seem to work. |
5edb9387 | 73 | 2.13. I have killed ypbind to stop using NIS, but glibc |
3dcf8ea6 | 74 | continues using NIS. |
5edb9387 | 75 | 2.14. Under Linux/Alpha, I always get "do_ypcall: clnt_call: |
3dcf8ea6 | 76 | RPC: Unable to receive; errno = Connection refused" when using NIS. |
5edb9387 UD |
77 | 2.15. After installing glibc name resolving doesn't work properly. |
78 | 2.16. How do I create the databases for NSS? | |
79 | 2.17. I have /usr/include/net and /usr/include/scsi as symlinks | |
61952351 | 80 | into my Linux source tree. Is that wrong? |
5edb9387 | 81 | 2.18. Programs like `logname', `top', `uptime' `users', `w' and |
61952351 UD |
82 | `who', show incorrect information about the (number of) |
83 | users on my system. Why? | |
5edb9387 | 84 | 2.19. After upgrading to glibc 2.1 with symbol versioning I get |
61952351 | 85 | errors about undefined symbols. What went wrong? |
5edb9387 | 86 | 2.20. When I start the program XXX after upgrading the library |
61952351 UD |
87 | I get |
88 | XXX: Symbol `_sys_errlist' has different size in shared | |
89 | object, consider re-linking | |
90 | Why? What should I do? | |
5edb9387 UD |
91 | 2.21. What do I need for C++ development? |
92 | 2.22. Even statically linked programs need some shared libraries | |
ff44f2a5 | 93 | which is not acceptable for me. What can I do? |
5edb9387 | 94 | 2.23. I just upgraded my Linux system to glibc and now I get |
fdacb17d | 95 | errors whenever I try to link any program. |
5edb9387 | 96 | 2.24. When I use nscd the machine freezes. |
0155a773 | 97 | 2.25. I need lots of open files. What do I have to do? |
7db169c9 UD |
98 | 2.26. How do I get the same behavior on parsing /etc/passwd and |
99 | /etc/group as I have with libc5 ? | |
b710a6e2 UD |
100 | 2.27. What needs to be recompiled when upgrading from glibc 2.0 to glibc |
101 | 2.1? | |
b7398be5 | 102 | 2.28. Why is extracting files via tar so slow? |
2ee511d9 UD |
103 | 2.29. Compiling programs I get parse errors in libio.h (e.g. "parse error |
104 | before `_IO_seekoff'"). How should I fix this? | |
4f7ea427 UD |
105 | 2.30. After upgrading to glibc 2.1, libraries that were compiled against |
106 | glibc 2.0.x don't work anymore. | |
61952351 UD |
107 | |
108 | 3. Source and binary incompatibilities, and what to do about them | |
109 | ||
110 | 3.1. I expect GNU libc to be 100% source code compatible with | |
111 | the old Linux based GNU libc. Why isn't it like this? | |
112 | 3.2. Why does getlogin() always return NULL on my Linux box? | |
113 | 3.3. Where are the DST_* constants found in <sys/time.h> on many | |
114 | systems? | |
115 | 3.4. The prototypes for `connect', `accept', `getsockopt', | |
116 | `setsockopt', `getsockname', `getpeername', `send', | |
117 | `sendto', and `recvfrom' are different in GNU libc from | |
118 | any other system I saw. This is a bug, isn't it? | |
119 | 3.5. On Linux I've got problems with the declarations in Linux | |
120 | kernel headers. | |
121 | 3.6. I don't include any kernel headers myself but the compiler | |
122 | still complains about redeclarations of types in the kernel | |
123 | headers. | |
124 | 3.7. Why don't signals interrupt system calls anymore? | |
125 | 3.8. I've got errors compiling code that uses certain string | |
126 | functions. Why? | |
4775243a UD |
127 | 3.9. I get compiler messages "Initializer element not constant" with |
128 | stdin/stdout/stderr. Why? | |
129 | 3.10. I can't compile with gcc -traditional (or | |
130 | -traditional-cpp). Why? | |
131 | 3.11. I get some errors with `gcc -ansi'. Isn't glibc ANSI compatible? | |
a35cb74d UD |
132 | 3.12. I can't access some functions anymore. nm shows that they do |
133 | exist but linking fails nevertheless. | |
a5f4e34a UD |
134 | 3.13. When using the db-2 library which comes with glibc is used in |
135 | the Perl db modules the testsuite is not passed. This did not | |
136 | happen with db-1, gdbm, or ndbm. | |
5148d49f UD |
137 | 3.14. The pow() inline function I get when including <math.h> is broken. |
138 | I get segmentation faults when I run the program. | |
05f732b3 | 139 | 3.15. The sys/sem.h file lacks the definition of `union semun'. |
33127459 | 140 | 3.16. Why has <netinet/ip_fw.h> disappeared? |
28ab8526 UD |
141 | 3.17. I get floods of warnings when I use -Wconversion and include |
142 | <string.h> or <math.h>. | |
5ff1a70a UD |
143 | 3.18. After upgrading to glibc 2.1, I receive errors about |
144 | unresolved symbols, like `_dl_initial_searchlist' and can not | |
145 | execute any binaries. What went wrong? | |
7d1de115 UD |
146 | 3.19. bonnie reports that char i/o with glibc 2 is much slower than with |
147 | libc5. What can be done? | |
b93492aa UD |
148 | 3.20. Programs compiled with glibc 2.1 can't read db files made with glibc |
149 | 2.0. What has changed that programs like rpm break? | |
b5a9efcd UD |
150 | 3.21. Autoconf's AC_CHECK_FUNC macro reports that a function exists, but |
151 | when I try to use it, it always returns -1 and sets errno to ENOSYS. | |
152 | 3.22. My program segfaults when I call fclose() on the FILE* returned | |
153 | from setmntent(). Is this a glibc bug? | |
61952351 | 154 | |
7d1de115 | 155 | 4. Miscellaneous |
61952351 | 156 | |
7d1de115 | 157 | 4.1. After I changed configure.in I get `Autoconf version X.Y. |
61952351 | 158 | or higher is required for this script'. What can I do? |
7d1de115 | 159 | 4.2. When I try to compile code which uses IPv6 headers and |
61952351 UD |
160 | definitions on my Linux 2.x.y system I am in trouble. |
161 | Nothing seems to work. | |
7d1de115 | 162 | 4.3. When I set the timezone by setting the TZ environment variable |
ff44f2a5 UD |
163 | to EST5EDT things go wrong since glibc computes the wrong time |
164 | from this information. | |
7d1de115 UD |
165 | 4.4. What other sources of documentation about glibc are available? |
166 | 4.5. The timezone string for Sydney/Australia is wrong since even when | |
348ed515 | 167 | daylight saving time is in effect the timezone string is EST. |
7d1de115 | 168 | 4.6. I've build make 3.77 against glibc 2.1 and now make gets |
eeabe877 | 169 | segmentation faults. |
f8cac037 | 170 | |
61952351 UD |
171 | \f |
172 | ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ | |
f4017d20 | 173 | |
61952351 | 174 | 1. Compiling glibc |
04be94a8 | 175 | |
61952351 | 176 | 1.1. What systems does the GNU C Library run on? |
613a76ff | 177 | |
f12944ec UD |
178 | {UD} This is difficult to answer. The file `README' lists the architectures |
179 | GNU libc was known to run on *at some time*. This does not mean that it | |
180 | still can be compiled and run on them now. | |
f8cac037 | 181 | |
f12944ec UD |
182 | The systems glibc is known to work on as of this release, and most probably |
183 | in the future, are: | |
f8cac037 RM |
184 | |
185 | *-*-gnu GNU Hurd | |
4775243a UD |
186 | i[3456]86-*-linux-gnu Linux-2.x on Intel |
187 | m68k-*-linux-gnu Linux-2.x on Motorola 680x0 | |
188 | alpha-*-linux-gnu Linux-2.x on DEC Alpha | |
9a0a462c | 189 | powerpc-*-linux-gnu Linux and MkLinux on PowerPC systems |
4775243a UD |
190 | sparc-*-linux-gnu Linux-2.x on SPARC |
191 | sparc64-*-linux-gnu Linux-2.x on UltraSPARC | |
ff44f2a5 | 192 | arm-*-none ARM standalone systems |
348ed515 | 193 | arm-*-linux Linux-2.x on ARM |
ff44f2a5 | 194 | arm-*-linuxaout Linux-2.x on ARM using a.out binaries |
f8cac037 | 195 | |
f12944ec UD |
196 | Ports to other Linux platforms are in development, and may in fact work |
197 | already, but no one has sent us success reports for them. Currently no | |
198 | ports to other operating systems are underway, although a few people have | |
199 | expressed interest. | |
f8cac037 | 200 | |
f12944ec UD |
201 | If you have a system not listed above (or in the `README' file) and you are |
202 | really interested in porting it, contact | |
f8cac037 | 203 | |
4775243a | 204 | <bug-glibc@gnu.org> |
f8cac037 RM |
205 | |
206 | ||
61952351 | 207 | 1.2. What compiler do I need to build GNU libc? |
f8cac037 | 208 | |
f12944ec UD |
209 | {UD} You must use GNU CC to compile GNU libc. A lot of extensions of GNU CC |
210 | are used to increase portability and speed. | |
f8cac037 | 211 | |
61952351 | 212 | GNU CC is found, like all other GNU packages, on |
f12944ec | 213 | |
a35cb74d | 214 | ftp://ftp.gnu.org/pub/gnu |
f12944ec | 215 | |
a35cb74d | 216 | and the many mirror sites. ftp.gnu.org is always overloaded, so try to find |
61952351 | 217 | a local mirror first. |
f8cac037 | 218 | |
b0610668 | 219 | You should always try to use the latest official release. Older versions |
f12944ec | 220 | may not have all the features GNU libc requires. The current releases of |
b8f558b7 UD |
221 | egcs (1.0.3 and 1.1.1) should work with the GNU C library (for powerpc see |
222 | question 1.5; for ARM see question 1.6). | |
f8cac037 | 223 | |
95f7cecb UD |
224 | While the GNU CC should be able to compile glibc it is nevertheless adviced |
225 | to use EGCS. Comparing the sizes of glibc on Intel compiled with a recent | |
226 | EGCS and gcc 2.8.1 shows this: | |
227 | ||
228 | text data bss dec hex filename | |
66f6a52b UD |
229 | egcs-2.93.10 862897 15944 12824 891665 d9b11 libc.so |
230 | gcc-2.8.1 959965 16468 12152 988585 f15a9 libc.so | |
95f7cecb UD |
231 | |
232 | Make up your own decision. | |
5edb9387 | 233 | |
f8cac037 | 234 | |
61952351 UD |
235 | 1.3. When I try to compile glibc I get only error messages. |
236 | What's wrong? | |
f8cac037 | 237 | |
b1418d8f | 238 | {UD} You definitely need GNU make to build GNU libc. No other make |
f12944ec | 239 | program has the needed functionality. |
f8cac037 | 240 | |
5edb9387 UD |
241 | We recommend version GNU make version 3.75 or 3.77. Versions before 3.75 |
242 | have bugs and/or are missing features. Version 3.76 has bugs which | |
243 | appear when building big projects like GNU libc. 3.76.1 appears to work but | |
7db169c9 | 244 | some people have reported problems. If you build GNU make 3.77 from source, |
7d1de115 | 245 | please read question 4.6 first. |
f8cac037 | 246 | |
f8cac037 | 247 | |
5edb9387 | 248 | 1.4. Do I need a special linker or assembler? |
f8cac037 | 249 | |
5edb9387 UD |
250 | {ZW} If you want a shared library, you need a linker and assembler that |
251 | understand all the features of ELF, including weak and versioned symbols. | |
252 | The static library can be compiled with less featureful tools, but lacks key | |
253 | features such as NSS. | |
41f27456 | 254 | |
5edb9387 UD |
255 | For Linux or Hurd, you want binutils 2.8.1.0.23, 2.9.1, or 2.9.1.0.15 or |
256 | higher. These are the only versions we've tested and found reliable. Other | |
257 | versions after 2.8.1.0.23 may work but we don't recommend them, especially | |
258 | not when C++ is involved. Earlier versions do not work at all. | |
a379e56a | 259 | |
5edb9387 UD |
260 | Other operating systems may come with system tools that have all the |
261 | necessary features, but this is moot because glibc hasn't been ported to | |
262 | them. | |
f8cac037 | 263 | |
f8cac037 | 264 | |
8619129f | 265 | 1.5. Which compiler should I use for powerpc? |
4775243a | 266 | |
48244d09 | 267 | {GK} You want to use egcs 1.1 or later (together with the right versions |
f12944ec | 268 | of all the other tools, of course). |
4775243a | 269 | |
48244d09 UD |
270 | In fact, egcs 1.1 has a bug that causes linuxthreads to be |
271 | miscompiled, resulting in segmentation faults when using condition | |
272 | variables. There is a temporary patch at: | |
4775243a | 273 | |
48244d09 | 274 | <http://discus.anu.edu.au/~geoffk/egcs-3.diff> |
4775243a | 275 | |
48244d09 | 276 | Later versions of egcs may fix this problem. |
4775243a UD |
277 | |
278 | ||
348ed515 UD |
279 | 1.6. Which tools should I use for ARM? |
280 | ||
281 | {PB} You should use egcs 1.1 or a later version. For ELF systems some | |
282 | changes are needed to the compiler; a patch against egcs-1.1.x can be found | |
283 | at: | |
284 | ||
285 | <ftp://ftp.netwinder.org/users/p/philb/egcs-1.1.1pre2-diff-981126> | |
286 | ||
287 | Binutils 2.9.1.0.16 or later is also required. | |
288 | ||
289 | ||
290 | 1.7. Do I need some more things to compile the GNU C Library? | |
f8cac037 | 291 | |
61952351 | 292 | {UD} Yes, there are some more :-). |
78b5ba3e | 293 | |
61952351 UD |
294 | * GNU gettext. This package contains the tools needed to construct |
295 | `message catalog' files containing translated versions of system | |
a35cb74d | 296 | messages. See ftp://ftp.gnu.org/pub/gnu or better any mirror |
61952351 | 297 | site. (We distribute compiled message catalogs, but they may not be |
2f2f52f7 UD |
298 | updated in patches.) Please note that the required minimal version |
299 | (0.10.35) of gettext is alpha software and available from | |
300 | ftp://alpha.gnu.org/gnu . | |
f8cac037 | 301 | |
5edb9387 UD |
302 | * Some files are built with special tools. E.g., files ending in .gperf |
303 | need a `gperf' program. The GNU version (now available in a separate | |
304 | package, formerly only as part of libg++) is known to work while some | |
305 | vendor versions do not. | |
f8cac037 | 306 | |
61952351 | 307 | You should not need these tools unless you change the source files. |
1f205a47 | 308 | |
5edb9387 UD |
309 | * Perl 5 is needed if you wish to test an installation of GNU libc |
310 | as the primary C library. | |
4775243a | 311 | |
61952351 UD |
312 | * When compiling for Linux, the header files of the Linux kernel must |
313 | be available to the compiler as <linux/*.h> and <asm/*.h>. | |
f8cac037 | 314 | |
8619129f UD |
315 | * lots of disk space (~170MB for i?86-linux; more for RISC platforms, |
316 | as much as 400MB). | |
af6f3906 | 317 | |
61952351 | 318 | * plenty of time. Compiling just the shared and static libraries for |
b5a9efcd UD |
319 | i?86-linux takes approximately 1h on an AMD-K6@225MHz w/ 96MB of RAM, |
320 | 45mins on a Celeron@400MHz w/ 128MB, and 55mins on a Alpha@533MHz w/ 256MB. | |
321 | Multiply this by 1.5 or 2.0 if you build profiling and/or the highly | |
322 | optimized version as well. For Hurd systems times are much higher. | |
f8cac037 | 323 | |
61952351 UD |
324 | You should avoid compiling in a NFS mounted filesystem. This is |
325 | very slow. | |
0200214b | 326 | |
61952351 | 327 | James Troup <J.J.Troup@comp.brad.ac.uk> reports a compile time of |
4775243a UD |
328 | 45h34m for a full build (shared, static, and profiled) on Atari |
329 | Falcon (Motorola 68030 @ 16 Mhz, 14 Mb memory) and Jan Barte | |
330 | <yann@plato.uni-paderborn.de> reports 22h48m on Atari TT030 | |
331 | (Motorola 68030 @ 32 Mhz, 34 Mb memory) | |
0200214b | 332 | |
61952351 | 333 | If you have some more measurements let me know. |
0200214b | 334 | |
ba1ffaa1 | 335 | |
348ed515 | 336 | 1.8. What version of the Linux kernel headers should be used? |
a35cb74d | 337 | |
f12944ec UD |
338 | {AJ,UD} The headers from the most recent Linux kernel should be used. The |
339 | headers used while compiling the GNU C library and the kernel binary used | |
340 | when using the library do not need to match. The GNU C library runs without | |
341 | problems on kernels that are older than the kernel headers used. The other | |
342 | way round (compiling the GNU C library with old kernel headers and running | |
343 | on a recent kernel) does not necessarily work. For example you can't use | |
b1418d8f | 344 | new kernel features if you used old kernel headers to compile the GNU C |
f12944ec UD |
345 | library. |
346 | ||
b0610668 | 347 | {ZW} Even if you are using a 2.0 kernel on your machine, we recommend you |
b710a6e2 UD |
348 | compile GNU libc with 2.2 kernel headers. That way you won't have to |
349 | recompile libc if you ever upgrade to kernel 2.2. To tell libc which | |
b0610668 | 350 | headers to use, give configure the --with-headers switch |
b710a6e2 | 351 | (e.g. --with-headers=/usr/src/linux-2.2.0/include). |
b0610668 | 352 | |
b710a6e2 | 353 | Note that you must configure the 2.2 kernel if you do this, otherwise libc |
62595351 | 354 | will be unable to find <linux/version.h>. Just change the current directory |
b710a6e2 | 355 | to the root of the 2.2 tree and do `make include/linux/version.h'. |
b0610668 | 356 | |
f12944ec | 357 | |
348ed515 | 358 | 1.9. The compiler hangs while building iconvdata modules. What's |
f12944ec UD |
359 | wrong? |
360 | ||
5edb9387 UD |
361 | {ZW} This is a problem with old versions of GCC. Initialization of large |
362 | static arrays is very slow. The compiler will eventually finish; give it | |
363 | time. | |
a35cb74d | 364 | |
b8f558b7 | 365 | The problem is fixed in egcs 1.1. |
a35cb74d | 366 | |
f12944ec | 367 | |
348ed515 | 368 | 1.10. When I run `nm -u libc.so' on the produced library I still |
61952351 | 369 | find unresolved symbols. Can this be ok? |
f8cac037 | 370 | |
f12944ec | 371 | {UD} Yes, this is ok. There can be several kinds of unresolved symbols: |
f8cac037 | 372 | |
61952351 UD |
373 | * magic symbols automatically generated by the linker. These have names |
374 | like __start_* and __stop_* | |
f8cac037 | 375 | |
78b5ba3e RM |
376 | * symbols starting with _dl_* come from the dynamic linker |
377 | ||
61952351 | 378 | * weak symbols, which need not be resolved at all (fabs for example) |
f8cac037 RM |
379 | |
380 | Generally, you should make sure you find a real program which produces | |
41f27456 | 381 | errors while linking before deciding there is a problem. |
f8cac037 RM |
382 | |
383 | ||
348ed515 | 384 | 1.11. What are these `add-ons'? |
999493cb | 385 | |
f12944ec UD |
386 | {UD} To avoid complications with export rules or external source code some |
387 | optional parts of the libc are distributed as separate packages (e.g., the | |
388 | crypt package, see question 2.5). | |
999493cb | 389 | |
f12944ec UD |
390 | To use these packages as part of GNU libc, just unpack the tarfiles in the |
391 | libc source directory and tell the configuration script about them using the | |
392 | --enable-add-ons option. If you give just --enable-add-ons configure tries | |
393 | to find all the add-on packages in your source tree. This may not work. If | |
394 | it doesn't, or if you want to select only a subset of the add-ons, give a | |
395 | comma-separated list of the add-ons to enable: | |
613a76ff | 396 | |
61952351 | 397 | configure --enable-add-ons=crypt,linuxthreads |
41f27456 | 398 | |
61952351 | 399 | for example. |
0200214b | 400 | |
f12944ec UD |
401 | Add-ons can add features (including entirely new shared libraries), override |
402 | files, provide support for additional architectures, and just about anything | |
403 | else. The existing makefiles do most of the work; only some few stub rules | |
404 | must be written to get everything running. | |
613a76ff | 405 | |
613a76ff | 406 | |
348ed515 | 407 | 1.12. My XXX kernel emulates a floating-point coprocessor for me. |
61952351 | 408 | Should I enable --with-fp? |
613a76ff | 409 | |
f12944ec UD |
410 | {ZW} An emulated FPU is just as good as a real one, as far as the C library |
411 | is concerned. You only need to say --without-fp if your machine has no way | |
412 | to execute floating-point instructions. | |
f8cac037 | 413 | |
61952351 UD |
414 | People who are interested in squeezing the last drop of performance |
415 | out of their machine may wish to avoid the trap overhead, but this is | |
416 | far more trouble than it's worth: you then have to compile | |
417 | *everything* this way, including the compiler's internal libraries | |
418 | (libgcc.a for GNU C), because the calling conventions change. | |
a1470b6f | 419 | |
999493cb | 420 | |
348ed515 | 421 | 1.13. When compiling GNU libc I get lots of errors saying functions |
61952351 | 422 | in glibc are duplicated in libgcc. |
5290baf0 | 423 | |
f12944ec UD |
424 | {EY} This is *exactly* the same problem that I was having. The problem was |
425 | due to the fact that configure didn't correctly detect that the linker flag | |
426 | --no-whole-archive was supported in my linker. In my case it was because I | |
427 | had run ./configure with bogus CFLAGS, and the test failed. | |
78b5ba3e | 428 | |
f12944ec UD |
429 | One thing that is particularly annoying about this problem is that once this |
430 | is misdetected, running configure again won't fix it unless you first delete | |
431 | config.cache. | |
78b5ba3e | 432 | |
f12944ec UD |
433 | {UD} Starting with glibc-2.0.3 there should be a better test to avoid some |
434 | problems of this kind. The setting of CFLAGS is checked at the very | |
435 | beginning and if it is not usable `configure' will bark. | |
78b5ba3e | 436 | |
af6f3906 | 437 | |
348ed515 | 438 | 1.14. Why do I get messages about missing thread functions when I use |
a35cb74d | 439 | librt? I don't even use threads. |
4775243a | 440 | |
a35cb74d UD |
441 | {UD} In this case you probably mixed up your installation. librt uses |
442 | threads internally and has implicit references to the thread library. | |
f12944ec UD |
443 | Normally these references are satisfied automatically but if the thread |
444 | library is not in the expected place you must tell the linker where it is. | |
445 | When using GNU ld it works like this: | |
4775243a UD |
446 | |
447 | gcc -o foo foo.c -Wl,-rpath-link=/some/other/dir -lrt | |
448 | ||
f12944ec UD |
449 | The `/some/other/dir' should contain the thread library. `ld' will use the |
450 | given path to find the implicitly referenced library while not disturbing | |
451 | any other link path. | |
4775243a UD |
452 | |
453 | ||
348ed515 | 454 | 1.15. What's the problem with configure --enable-omitfp? |
78b5ba3e | 455 | |
61952351 | 456 | {AJ} When --enable-omitfp is set the libraries are built without frame |
fdacb17d | 457 | pointers. Some compilers produce buggy code for this model and therefore we |
f12944ec | 458 | don't advise using it at the moment. |
66219c07 | 459 | |
fdacb17d | 460 | If you use --enable-omitfp, you're on your own. If you encounter problems |
f12944ec UD |
461 | with a library that was build this way, we advise you to rebuild the library |
462 | without --enable-omitfp. If the problem vanishes consider tracking the | |
463 | problem down and report it as compiler failure. | |
66219c07 | 464 | |
b1418d8f UD |
465 | Since a library built with --enable-omitfp is undebuggable on most systems, |
466 | debuggable libraries are also built - you can use them by appending "_g" to | |
f12944ec | 467 | the library names. |
66219c07 | 468 | |
f12944ec UD |
469 | The compilation of these extra libraries and the compiler optimizations slow |
470 | down the build process and need more disk space. | |
66219c07 | 471 | |
b0610668 | 472 | |
b1418d8f | 473 | 1.16. I get failures during `make check'. What should I do? |
b0610668 | 474 | |
b1418d8f UD |
475 | {AJ} The testsuite should compile and run cleanly on your system; every |
476 | failure should be looked into. Depending on the failures, you probably | |
477 | should not install the library at all. | |
b0610668 UD |
478 | |
479 | You should consider using the `glibcbug' script to report the failure, | |
480 | providing as much detail as possible. If you run a test directly, please | |
481 | remember to set up the environment correctly. You want to test the compiled | |
482 | library - and not your installed one. The best way is to copy the exact | |
483 | command line which failed and run the test from the subdirectory for this | |
484 | test in the sources. | |
485 | ||
486 | There are some failures which are not directly related to the GNU libc: | |
b1418d8f UD |
487 | - Some compilers produce buggy code. No compiler gets single precision |
488 | complex numbers correct on Alpha. Otherwise, the egcs 1.1 release should be | |
489 | ok; gcc 2.8.1 might cause some failures; gcc 2.7.2.x is so buggy that | |
490 | explicit checks have been used so that you can't build with it. | |
b0610668 UD |
491 | - The kernel might have bugs. For example on Linux/Alpha 2.0.34 the |
492 | floating point handling has quite a number of bugs and therefore most of | |
b710a6e2 | 493 | the test cases in the math subdirectory will fail. Linux 2.2 has |
b1418d8f UD |
494 | fixes for the floating point support on Alpha. The Linux/SPARC kernel has |
495 | also some bugs in the FPU emulation code (as of Linux 2.2.0). | |
d32a4020 UD |
496 | - Other tools might have problems. For example bash 2.03 gives a |
497 | segmentation fault running the tst-rpmatch.sh test script. | |
b0610668 | 498 | |
a379e56a | 499 | |
348ed515 | 500 | 1.17. What is symbol versioning good for? Do I need it? |
a379e56a UD |
501 | |
502 | {AJ} Symbol versioning solves problems that are related to interface | |
503 | changes. One version of an interface might have been introduced in a | |
504 | previous version of the GNU C library but the interface or the semantics of | |
505 | the function has been changed in the meantime. For binary compatibility | |
506 | with the old library, a newer library needs to still have the old interface | |
b1418d8f | 507 | for old programs. On the other hand, new programs should use the new |
a379e56a | 508 | interface. Symbol versioning is the solution for this problem. The GNU |
b1418d8f UD |
509 | libc version 2.1 uses symbol versioning by default if the installed binutils |
510 | supports it. | |
a379e56a | 511 | |
b1418d8f UD |
512 | We don't advise building without symbol versioning, since you lose binary |
513 | compatibility - forever! The binary compatibility you lose is not only | |
514 | against the previous version of the GNU libc (version 2.0) but also against | |
515 | all future versions. | |
a379e56a | 516 | |
61952351 UD |
517 | \f |
518 | . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . | |
e6c9a67a | 519 | |
61952351 | 520 | 2. Installation and configuration issues |
e6c9a67a | 521 | |
61952351 | 522 | 2.1. Can I replace the libc on my Linux system with GNU libc? |
e6c9a67a | 523 | |
f12944ec UD |
524 | {UD} You cannot replace any existing libc for Linux with GNU libc. It is |
525 | binary incompatible and therefore has a different major version. You can, | |
526 | however, install it alongside your existing libc. | |
e6c9a67a | 527 | |
61952351 UD |
528 | For Linux there are three major libc versions: |
529 | libc-4 a.out libc | |
530 | libc-5 original ELF libc | |
531 | libc-6 GNU libc | |
e6c9a67a | 532 | |
f12944ec UD |
533 | You can have any combination of these three installed. For more information |
534 | consult documentation for shared library handling. The Makefiles of GNU | |
535 | libc will automatically generate the needed symbolic links which the linker | |
536 | will use. | |
e6c9a67a RM |
537 | |
538 | ||
61952351 UD |
539 | 2.2. How do I configure GNU libc so that the essential libraries |
540 | like libc.so go into /lib and the other into /usr/lib? | |
ec42724d | 541 | |
61952351 UD |
542 | {UD,AJ} Like all other GNU packages GNU libc is designed to use a base |
543 | directory and install all files relative to this. The default is | |
f12944ec UD |
544 | /usr/local, because this is safe (it will not damage the system if installed |
545 | there). If you wish to install GNU libc as the primary C library on your | |
546 | system, set the base directory to /usr (i.e. run configure --prefix=/usr | |
547 | <other_options>). Note that this can damage your system; see question 2.3 for | |
548 | details. | |
549 | ||
550 | Some systems like Linux have a filesystem standard which makes a difference | |
551 | between essential libraries and others. Essential libraries are placed in | |
552 | /lib because this directory is required to be located on the same disk | |
553 | partition as /. The /usr subtree might be found on another | |
554 | partition/disk. If you configure for Linux with --prefix=/usr, then this | |
555 | will be done automatically. | |
ec42724d | 556 | |
61952351 | 557 | To install the essential libraries which come with GNU libc in /lib on |
f12944ec UD |
558 | systems other than Linux one must explicitly request it. Autoconf has no |
559 | option for this so you have to use a `configparms' file (see the `INSTALL' | |
560 | file for details). It should contain: | |
ec42724d RM |
561 | |
562 | slibdir=/lib | |
563 | sysconfdir=/etc | |
564 | ||
f12944ec UD |
565 | The first line specifies the directory for the essential libraries, the |
566 | second line the directory for system configuration files. | |
ec42724d | 567 | |
5290baf0 | 568 | |
61952351 | 569 | 2.3. How should I avoid damaging my system when I install GNU libc? |
ec42724d | 570 | |
f12944ec UD |
571 | {ZW} If you wish to be cautious, do not configure with --prefix=/usr. If |
572 | you don't specify a prefix, glibc will be installed in /usr/local, where it | |
573 | will probably not break anything. (If you wish to be certain, set the | |
574 | prefix to something like /usr/local/glibc2 which is not used for anything.) | |
845dcb57 | 575 | |
61952351 | 576 | The dangers when installing glibc in /usr are twofold: |
845dcb57 | 577 | |
61952351 | 578 | * glibc will overwrite the headers in /usr/include. Other C libraries |
27e309c1 UD |
579 | install a different but overlapping set of headers there, so the effect |
580 | will probably be that you can't compile anything. You need to rename | |
581 | /usr/include out of the way before running `make install'. (Do not throw | |
582 | it away; you will then lose the ability to compile programs against your | |
583 | old libc.) | |
845dcb57 | 584 | |
61952351 UD |
585 | * None of your old libraries, static or shared, can be used with a |
586 | different C library major version. For shared libraries this is not a | |
587 | problem, because the filenames are different and the dynamic linker | |
588 | will enforce the restriction. But static libraries have no version | |
589 | information. You have to evacuate all the static libraries in | |
590 | /usr/lib to a safe location. | |
845dcb57 | 591 | |
61952351 UD |
592 | The situation is rather similar to the move from a.out to ELF which |
593 | long-time Linux users will remember. | |
845dcb57 | 594 | |
845dcb57 | 595 | |
61952351 UD |
596 | 2.4. Do I need to use GNU CC to compile programs that will use the |
597 | GNU C Library? | |
845dcb57 | 598 | |
f12944ec UD |
599 | {ZW} In theory, no; the linker does not care, and the headers are supposed |
600 | to check for GNU CC before using its extensions to the C language. | |
845dcb57 | 601 | |
f12944ec UD |
602 | However, there are currently no ports of glibc to systems where another |
603 | compiler is the default, so no one has tested the headers extensively | |
604 | against another compiler. You may therefore encounter difficulties. If you | |
605 | do, please report them as bugs. | |
845dcb57 | 606 | |
61952351 UD |
607 | Also, in several places GNU extensions provide large benefits in code |
608 | quality. For example, the library has hand-optimized, inline assembly | |
f12944ec UD |
609 | versions of some string functions. These can only be used with GCC. See |
610 | question 3.8 for details. | |
845dcb57 | 611 | |
845dcb57 | 612 | |
61952351 UD |
613 | 2.5. When linking with the new libc I get unresolved symbols |
614 | `crypt' and `setkey'. Why aren't these functions in the | |
615 | libc anymore? | |
845dcb57 | 616 | |
f12944ec UD |
617 | {UD} The US places restrictions on exporting cryptographic programs and |
618 | source code. Until this law gets abolished we cannot ship the cryptographic | |
619 | functions together with glibc. | |
845dcb57 | 620 | |
348ed515 | 621 | The functions are available, as an add-on (see question 1.11). People in the US |
f12944ec | 622 | may get it from the same place they got GNU libc from. People outside the |
9f2a9248 AS |
623 | US should get the code from ftp.funet.fi [128.214.248.6] in the directory |
624 | pub/gnu/funet, or another archive site outside the USA. The README explains | |
625 | how to install the sources. | |
c4029823 | 626 | |
f12944ec UD |
627 | If you already have the crypt code on your system the reason for the failure |
628 | is probably that you did not link with -lcrypt. The crypto functions are in | |
629 | a separate library to make it possible to export GNU libc binaries from the | |
630 | US. | |
c4029823 | 631 | |
c4029823 | 632 | |
61952351 UD |
633 | 2.6. When I use GNU libc on my Linux system by linking against |
634 | the libc.so which comes with glibc all I get is a core dump. | |
c4029823 | 635 | |
f12944ec UD |
636 | {UD} On Linux, gcc sets the dynamic linker to /lib/ld-linux.so.1 unless the |
637 | user specifies a -dynamic-linker argument. This is the name of the libc5 | |
638 | dynamic linker, which does not work with glibc. | |
61952351 | 639 | |
a379e56a UD |
640 | For casual use of GNU libc you can just specify to the linker |
641 | --dynamic-linker=/lib/ld-linux.so.2 | |
61952351 | 642 | |
f12944ec | 643 | which is the glibc dynamic linker, on Linux systems. On other systems the |
a379e56a UD |
644 | name is /lib/ld.so.1. When linking via gcc, you've got to add |
645 | -Wl,--dynamic-linker=/lib/ld-linux.so.2 | |
646 | ||
647 | to the gcc command line. | |
c4029823 | 648 | |
f12944ec UD |
649 | To change your environment to use GNU libc for compiling you need to change |
650 | the `specs' file of your gcc. This file is normally found at | |
c4029823 UD |
651 | |
652 | /usr/lib/gcc-lib/<arch>/<version>/specs | |
653 | ||
654 | In this file you have to change a few things: | |
655 | ||
61952351 | 656 | - change `ld-linux.so.1' to `ld-linux.so.2' |
c4029823 UD |
657 | |
658 | - remove all expression `%{...:-lgmon}'; there is no libgmon in glibc | |
659 | ||
f4017d20 UD |
660 | - fix a minor bug by changing %{pipe:-} to %| |
661 | ||
f12944ec UD |
662 | Here is what the gcc-2.7.2 specs file should look like when GNU libc is |
663 | installed at /usr: | |
c4029823 UD |
664 | |
665 | ----------------------------------------------------------------------- | |
666 | *asm: | |
667 | %{V} %{v:%{!V:-V}} %{Qy:} %{!Qn:-Qy} %{n} %{T} %{Ym,*} %{Yd,*} %{Wa,*:%*} | |
668 | ||
669 | *asm_final: | |
f4017d20 | 670 | %| |
c4029823 UD |
671 | |
672 | *cpp: | |
68dbb3a6 | 673 | %{fPIC:-D__PIC__ -D__pic__} %{fpic:-D__PIC__ -D__pic__} %{!m386:-D__i486__} %{posix:-D_POSIX_SOURCE} %{pthread:-D_REENTRANT} |
c4029823 UD |
674 | |
675 | *cc1: | |
68dbb3a6 | 676 | %{profile:-p} |
c4029823 UD |
677 | |
678 | *cc1plus: | |
679 | ||
680 | ||
681 | *endfile: | |
68dbb3a6 | 682 | %{!shared:crtend.o%s} %{shared:crtendS.o%s} crtn.o%s |
c4029823 UD |
683 | |
684 | *link: | |
68dbb3a6 | 685 | -m elf_i386 %{shared:-shared} %{!shared: %{!ibcs: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker /lib/ld-linux.so.2}} %{static:-static}}} |
c4029823 UD |
686 | |
687 | *lib: | |
68dbb3a6 | 688 | %{!shared: %{pthread:-lpthread} %{profile:-lc_p} %{!profile: -lc}} |
c4029823 UD |
689 | |
690 | *libgcc: | |
68dbb3a6 | 691 | -lgcc |
c4029823 UD |
692 | |
693 | *startfile: | |
61952351 | 694 | %{!shared: %{pg:gcrt1.o%s} %{!pg:%{p:gcrt1.o%s} %{!p:%{profile:gcrt1.o%s} %{!profile:crt1.o%s}}}} crti.o%s %{!shared:crtbegin.o%s} %{shared:crtbeginS.o%s} |
c4029823 UD |
695 | |
696 | *switches_need_spaces: | |
697 | ||
698 | ||
699 | *signed_char: | |
700 | %{funsigned-char:-D__CHAR_UNSIGNED__} | |
701 | ||
702 | *predefines: | |
703 | -D__ELF__ -Dunix -Di386 -Dlinux -Asystem(unix) -Asystem(posix) -Acpu(i386) -Amachine(i386) | |
704 | ||
705 | *cross_compile: | |
706 | 0 | |
707 | ||
708 | *multilib: | |
709 | . ; | |
710 | ||
711 | ----------------------------------------------------------------------- | |
712 | ||
f12944ec UD |
713 | Things get a bit more complicated if you have GNU libc installed in some |
714 | other place than /usr, i.e., if you do not want to use it instead of the old | |
715 | libc. In this case the needed startup files and libraries are not found in | |
716 | the regular places. So the specs file must tell the compiler and linker | |
717 | exactly what to use. | |
0d204b0a | 718 | |
f41c8091 | 719 | Version 2.7.2.3 does and future versions of GCC will automatically |
0d8733c4 | 720 | provide the correct specs. |
c4029823 UD |
721 | |
722 | ||
61952351 UD |
723 | 2.7. Looking through the shared libc file I haven't found the |
724 | functions `stat', `lstat', `fstat', and `mknod' and while | |
725 | linking on my Linux system I get error messages. How is | |
726 | this supposed to work? | |
c4029823 | 727 | |
f12944ec UD |
728 | {RM} Believe it or not, stat and lstat (and fstat, and mknod) are supposed |
729 | to be undefined references in libc.so.6! Your problem is probably a missing | |
730 | or incorrect /usr/lib/libc.so file; note that this is a small text file now, | |
731 | not a symlink to libc.so.6. It should look something like this: | |
c4029823 | 732 | |
ff44f2a5 | 733 | GROUP ( libc.so.6 libc_nonshared.a ) |
1f205a47 | 734 | |
c4029823 | 735 | |
5edb9387 UD |
736 | 2.8. When I run an executable on one system which I compiled on |
737 | another, I get dynamic linker errors. Both systems have the same | |
738 | version of glibc installed. What's wrong? | |
739 | ||
740 | {ZW} Glibc on one of these systems was compiled with gcc 2.7 or 2.8, the | |
741 | other with egcs (any version). Egcs has functions in its internal | |
742 | `libgcc.a' to support exception handling with C++. They are linked into | |
743 | any program or dynamic library compiled with egcs, whether it needs them or | |
744 | not. Dynamic libraries then turn around and export those functions again | |
745 | unless special steps are taken to prevent them. | |
746 | ||
747 | When you link your program, it resolves its references to the exception | |
748 | functions to the ones exported accidentally by libc.so. That works fine as | |
749 | long as libc has those functions. On the other system, libc doesn't have | |
750 | those functions because it was compiled by gcc 2.8, and you get undefined | |
751 | symbol errors. The symbols in question are named things like | |
752 | `__register_frame_info'. | |
753 | ||
754 | For glibc 2.0, the workaround is to not compile libc with egcs. We've also | |
755 | incorporated a patch which should prevent the EH functions sneaking into | |
756 | libc. It doesn't matter what compiler you use to compile your program. | |
757 | ||
758 | For glibc 2.1, we've chosen to do it the other way around: libc.so | |
759 | explicitly provides the EH functions. This is to prevent other shared | |
95f7cecb UD |
760 | libraries from doing it. |
761 | ||
762 | {UD} Starting with glibc 2.1.1 you can compile glibc with gcc 2.8.1 or | |
763 | newer since we have explicitly add references to the functions causing the | |
764 | problem. But you nevertheless should use EGCS for other reasons | |
765 | (see question 1.2). | |
5edb9387 UD |
766 | |
767 | ||
768 | 2.9. How can I compile gcc 2.7.2.1 from the gcc source code using | |
61952351 | 769 | glibc 2.x? |
ba1ffaa1 | 770 | |
f12944ec | 771 | {AJ} There's only correct support for glibc 2.0.x in gcc 2.7.2.3 or later. |
95f7cecb | 772 | But you should get at least gcc 2.8.1 or egcs 1.1 (or later versions) |
f12944ec | 773 | instead. |
ba1ffaa1 UD |
774 | |
775 | ||
5edb9387 | 776 | 2.10. The `gencat' utility cannot process the catalog sources which |
61952351 | 777 | were used on my Linux libc5 based system. Why? |
47707456 | 778 | |
f12944ec UD |
779 | {UD} The `gencat' utility provided with glibc complies to the XPG standard. |
780 | The older Linux version did not obey the standard, so they are not | |
781 | compatible. | |
47707456 | 782 | |
61952351 | 783 | To ease the transition from the Linux version some of the non-standard |
f12944ec UD |
784 | features are also present in the `gencat' program of GNU libc. This mainly |
785 | includes the use of symbols for the message number and the automatic | |
61952351 UD |
786 | generation of header files which contain the needed #defines to map the |
787 | symbols to integers. | |
47707456 | 788 | |
f12944ec UD |
789 | Here is a simple SED script to convert at least some Linux specific catalog |
790 | files to the XPG4 form: | |
68dbb3a6 | 791 | |
61952351 UD |
792 | ----------------------------------------------------------------------- |
793 | # Change catalog source in Linux specific format to standard XPG format. | |
794 | # Ulrich Drepper <drepper@cygnus.com>, 1996. | |
795 | # | |
796 | /^\$ #/ { | |
797 | h | |
798 | s/\$ #\([^ ]*\).*/\1/ | |
799 | x | |
800 | s/\$ #[^ ]* *\(.*\)/\$ \1/ | |
801 | } | |
68dbb3a6 | 802 | |
61952351 UD |
803 | /^# / { |
804 | s/^# \(.*\)/\1/ | |
805 | G | |
806 | s/\(.*\)\n\(.*\)/\2 \1/ | |
807 | } | |
808 | ----------------------------------------------------------------------- | |
19361cb7 | 809 | |
19361cb7 | 810 | |
5edb9387 | 811 | 2.11. Programs using libc have their messages translated, but other |
a35cb74d UD |
812 | behavior is not localized (e.g. collating order); why? |
813 | ||
814 | {ZW} Translated messages are automatically installed, but the locale | |
f12944ec UD |
815 | database that controls other behaviors is not. You need to run localedef to |
816 | install this database, after you have run `make install'. For example, to | |
817 | set up the French Canadian locale, simply issue the command | |
a35cb74d UD |
818 | |
819 | localedef -i fr_CA -f ISO-8859-1 fr_CA | |
820 | ||
821 | Please see localedata/README in the source tree for further details. | |
822 | ||
823 | ||
5edb9387 | 824 | 2.12. I have set up /etc/nis.conf, and the Linux libc 5 with NYS |
61952351 | 825 | works great. But the glibc NIS+ doesn't seem to work. |
19361cb7 | 826 | |
f12944ec UD |
827 | {TK} The glibc NIS+ implementation uses a /var/nis/NIS_COLD_START file for |
828 | storing information about the NIS+ server and their public keys, because the | |
829 | nis.conf file does not contain all the necessary information. You have to | |
830 | copy a NIS_COLD_START file from a Solaris client (the NIS_COLD_START file is | |
831 | byte order independent) or generate it with nisinit from the nis-tools | |
832 | package; available at | |
833 | ||
612fdf25 | 834 | http://www.suse.de/~kukuk/linux/nisplus.html |
19361cb7 | 835 | |
68dbb3a6 | 836 | |
5edb9387 | 837 | 2.13. I have killed ypbind to stop using NIS, but glibc |
3dcf8ea6 | 838 | continues using NIS. |
4d06461a | 839 | |
f12944ec UD |
840 | {TK} For faster NIS lookups, glibc uses the /var/yp/binding/ files from |
841 | ypbind. ypbind 3.3 and older versions don't always remove these files, so | |
842 | glibc will continue to use them. Other BSD versions seem to work correctly. | |
843 | Until ypbind 3.4 is released, you can find a patch at | |
844 | ||
66f6a52b | 845 | <ftp://ftp.kernel.org/pub/linux/utils/net/NIS/ypbind-3.3-glibc4.diff.gz> |
a35cb74d | 846 | |
4d06461a | 847 | |
5edb9387 | 848 | 2.14. Under Linux/Alpha, I always get "do_ypcall: clnt_call: |
3dcf8ea6 | 849 | RPC: Unable to receive; errno = Connection refused" when using NIS. |
4d06461a | 850 | |
f12944ec UD |
851 | {TK} You need a ypbind version which is 64bit clean. Some versions are not |
852 | 64bit clean. A 64bit clean implementation is ypbind-mt. For ypbind 3.3, | |
853 | you need the patch from ftp.kernel.org (See the previous question). I don't | |
854 | know about other versions. | |
a35cb74d UD |
855 | |
856 | ||
5edb9387 | 857 | 2.15. After installing glibc name resolving doesn't work properly. |
68dbb3a6 | 858 | |
f12944ec UD |
859 | {AJ} You probably should read the manual section describing nsswitch.conf |
860 | (just type `info libc "NSS Configuration File"'). The NSS configuration | |
861 | file is usually the culprit. | |
22d57dd3 | 862 | |
22d57dd3 | 863 | |
5edb9387 | 864 | 2.16. How do I create the databases for NSS? |
3dcf8ea6 UD |
865 | |
866 | {AJ} If you have an entry "db" in /etc/nsswitch.conf you should also create | |
867 | the database files. The glibc sources contain a Makefile which does the | |
a379e56a | 868 | necessary conversion and calls to create those files. The file is |
3dcf8ea6 UD |
869 | `db-Makefile' in the subdirectory `nss' and you can call it with `make -f |
870 | db-Makefile'. Please note that not all services are capable of using a | |
871 | database. Currently passwd, group, ethers, protocol, rpc, services shadow | |
872 | and netgroup are implemented. | |
873 | ||
874 | ||
5edb9387 | 875 | 2.17. I have /usr/include/net and /usr/include/scsi as symlinks |
61952351 | 876 | into my Linux source tree. Is that wrong? |
22d57dd3 | 877 | |
f12944ec UD |
878 | {PB} This was necessary for libc5, but is not correct when using glibc. |
879 | Including the kernel header files directly in user programs usually does not | |
880 | work (see question 3.5). glibc provides its own <net/*> and <scsi/*> header | |
881 | files to replace them, and you may have to remove any symlink that you have | |
882 | in place before you install glibc. However, /usr/include/asm and | |
883 | /usr/include/linux should remain as they were. | |
22d57dd3 | 884 | |
22d57dd3 | 885 | |
5edb9387 | 886 | 2.18. Programs like `logname', `top', `uptime' `users', `w' and |
61952351 UD |
887 | `who', show incorrect information about the (number of) |
888 | users on my system. Why? | |
22d57dd3 | 889 | |
61952351 | 890 | {MK} See question 3.2. |
22d57dd3 | 891 | |
22d57dd3 | 892 | |
5edb9387 | 893 | 2.19. After upgrading to glibc 2.1 with symbol versioning I get |
61952351 | 894 | errors about undefined symbols. What went wrong? |
26dee9c4 | 895 | |
f12944ec UD |
896 | {AJ} The problem is caused either by wrong program code or tools. In the |
897 | versioned libc a lot of symbols are now local that were global symbols in | |
898 | previous versions. It seems that programs linked against older versions | |
899 | often accidentally used libc global variables -- something that should not | |
900 | happen. | |
26dee9c4 | 901 | |
f12944ec UD |
902 | The only way to fix this is to recompile your program. Sorry, that's the |
903 | price you might have to pay once for quite a number of advantages with | |
904 | symbol versioning. | |
26dee9c4 | 905 | |
26dee9c4 | 906 | |
5edb9387 | 907 | 2.20. When I start the program XXX after upgrading the library |
61952351 UD |
908 | I get |
909 | XXX: Symbol `_sys_errlist' has different size in shared | |
910 | object, consider re-linking | |
911 | Why? What should I do? | |
26dee9c4 | 912 | |
f12944ec UD |
913 | {UD} As the message says, relink the binary. The problem is that a few |
914 | symbols from the library can change in size and there is no way to avoid | |
915 | this. _sys_errlist is a good example. Occasionally there are new error | |
916 | numbers added to the kernel and this must be reflected at user level, | |
917 | breaking programs that refer to them directly. | |
a2b08ee5 | 918 | |
f12944ec UD |
919 | Such symbols should normally not be used at all. There are mechanisms to |
920 | avoid using them. In the case of _sys_errlist, there is the strerror() | |
921 | function which should _always_ be used instead. So the correct fix is to | |
922 | rewrite that part of the application. | |
a2b08ee5 | 923 | |
f12944ec UD |
924 | In some situations (especially when testing a new library release) it might |
925 | be possible that a symbol changed size when that should not have happened. | |
926 | So in case of doubt report such a warning message as a problem. | |
a2b08ee5 | 927 | |
a35cb74d | 928 | |
5edb9387 UD |
929 | 2.21. What do I need for C++ development? |
930 | ||
931 | {HJ,AJ} You need either egcs 1.1 which comes directly with libstdc++ or | |
932 | gcc-2.8.1 together with libstdc++ 2.8.1.1. egcs 1.1 has the better C++ | |
933 | support and works directly with glibc 2.1. If you use gcc-2.8.1 with | |
934 | libstdc++ 2.8.1.1, you need to modify libstdc++ a bit. A patch is available | |
935 | as: | |
66f6a52b | 936 | <ftp://alpha.gnu.org/gnu/libstdc++-2.8.1.1-glibc2.1-diff.gz> |
a35cb74d | 937 | |
5edb9387 UD |
938 | Please note that libg++ 2.7.2 (and the Linux Versions 2.7.2.x) doesn't work |
939 | very well with the GNU C library due to vtable thunks. If you're upgrading | |
940 | from glibc 2.0.x to 2.1 you have to recompile libstdc++ since the library | |
941 | compiled for 2.0 is not compatible due to the new Large File Support (LFS) | |
942 | in version 2.1. | |
fb98e2bf UD |
943 | |
944 | {UD} But since in the case of a shared libstdc++ the version numbers should | |
945 | be different existing programs will continue to work. | |
a35cb74d | 946 | |
ff44f2a5 | 947 | |
5edb9387 | 948 | 2.22. Even statically linked programs need some shared libraries |
ff44f2a5 UD |
949 | which is not acceptable for me. What can I do? |
950 | ||
f12944ec UD |
951 | {AJ} NSS (for details just type `info libc "Name Service Switch"') won't |
952 | work properly without shared libraries. NSS allows using different services | |
953 | (e.g. NIS, files, db, hesiod) by just changing one configuration file | |
954 | (/etc/nsswitch.conf) without relinking any programs. The only disadvantage | |
955 | is that now static libraries need to access shared libraries. This is | |
956 | handled transparently by the GNU C library. | |
ff44f2a5 | 957 | |
f12944ec UD |
958 | A solution is to configure glibc with --enable-static-nss. In this case you |
959 | can create a static binary that will use only the services dns and files | |
960 | (change /etc/nsswitch.conf for this). You need to link explicitly against | |
961 | all these services. For example: | |
ff44f2a5 UD |
962 | |
963 | gcc -static test-netdb.c -o test-netdb.c \ | |
964 | -lc -lnss_files -lnss_dns -lresolv | |
965 | ||
966 | The problem with this approach is that you've got to link every static | |
967 | program that uses NSS routines with all those libraries. | |
968 | ||
969 | {UD} In fact, one cannot say anymore that a libc compiled with this | |
970 | option is using NSS. There is no switch anymore. Therefore it is | |
971 | *highly* recommended *not* to use --enable-static-nss since this makes | |
972 | the behaviour of the programs on the system inconsistent. | |
973 | ||
fdacb17d | 974 | |
5edb9387 | 975 | 2.23. I just upgraded my Linux system to glibc and now I get |
fdacb17d UD |
976 | errors whenever I try to link any program. |
977 | ||
978 | {ZW} This happens when you have installed glibc as the primary C library but | |
979 | have stray symbolic links pointing at your old C library. If the first | |
980 | `libc.so' the linker finds is libc 5, it will use that. Your program | |
981 | expects to be linked with glibc, so the link fails. | |
982 | ||
983 | The most common case is that glibc put its `libc.so' in /usr/lib, but there | |
984 | was a `libc.so' from libc 5 in /lib, which gets searched first. To fix the | |
985 | problem, just delete /lib/libc.so. You may also need to delete other | |
986 | symbolic links in /lib, such as /lib/libm.so if it points to libm.so.5. | |
987 | ||
988 | {AJ} The perl script test-installation.pl which is run as last step during | |
989 | an installation of glibc that is configured with --prefix=/usr should help | |
990 | detect these situations. If the script reports problems, something is | |
991 | really screwed up. | |
992 | ||
48244d09 | 993 | |
5edb9387 | 994 | 2.24. When I use nscd the machine freezes. |
48244d09 | 995 | |
5edb9387 UD |
996 | {UD} You cannot use nscd with Linux 2.0.*. There is functionality missing |
997 | in the kernel and work-arounds are not suitable. Besides, some parts of the | |
998 | kernel are too buggy when it comes to using threads. | |
48244d09 | 999 | |
b710a6e2 | 1000 | If you need nscd, you have to use at least a 2.1 kernel. |
48244d09 UD |
1001 | |
1002 | Note that I have at this point no information about any other platform. | |
1003 | ||
0155a773 UD |
1004 | |
1005 | 2.25. I need lots of open files. What do I have to do? | |
1006 | ||
1007 | {AJ} This is at first a kernel issue. The kernel defines limits with | |
1008 | OPEN_MAX the number of simultaneous open files and with FD_SETSIZE the | |
1009 | number of used file descriptors. You need to change these values in your | |
1010 | kernel and recompile the kernel so that the kernel allows to use more open | |
1011 | files. You don't necessarily need to recompile the GNU C library since the | |
1012 | only place where OPEN_MAX and FD_SETSIZE is really needed in the library | |
1013 | itself is the size of fd_set which is used by select. | |
1014 | ||
1015 | The GNU C library is now (nearly) select free. This means it internally has | |
1016 | no limits imposed by the `fd_set' type. Instead almost all places where the | |
1017 | functionality is needed the `poll' function is used. | |
1018 | ||
1019 | If you increase the number of file descriptors in the kernel you don't need | |
1020 | to recompile the C library. The remaining select calls are in the RPC code. | |
1021 | If your RPC daemons don't need more than FD_SETSIZE file descriptors, you | |
1022 | don't need to change anything at all. | |
1023 | ||
1024 | {UD} You can always get the maximum number of file descriptors a process is | |
1025 | allowed to have open at any time using | |
1026 | ||
1027 | number = sysconf (_SC_OPEN_MAX); | |
1028 | ||
1029 | This will work even if the kernel limits change. | |
1030 | ||
7db169c9 UD |
1031 | |
1032 | 2.26. How do I get the same behavior on parsing /etc/passwd and | |
1033 | /etc/group as I have with libc5 ? | |
1034 | ||
1035 | {TK} The name switch setup in /etc/nsswitch.conf selected by most Linux | |
1036 | distributions does not support +/- and netgroup entries in the files like | |
1037 | /etc/passwd. Though this is the preferred setup some people might have | |
1038 | setups coming over from the libc5 days where it was the default to recognize | |
1039 | lines like this. To get back to the old behaviour one simply has to change | |
1040 | the rules for passwd, group, and shadow in the nsswitch.conf file as | |
1041 | follows: | |
1042 | ||
1043 | passwd: compat | |
1044 | group: compat | |
1045 | shadow: compat | |
1046 | ||
1047 | passwd_compat: nis | |
1048 | group_compat: nis | |
1049 | shadow_compat: nis | |
1050 | ||
b710a6e2 UD |
1051 | |
1052 | 2.27. What needs to be recompiled when upgrading from glibc 2.0 to glibc | |
1053 | 2.1? | |
1054 | ||
1055 | {AJ,CG} If you just upgrade the glibc from 2.0.x (x <= 7) to 2.1, binaries | |
1056 | that have been linked against glibc 2.0 will continue to work. | |
1057 | ||
1058 | If you compile your own binaries against glibc 2.1, you also need to | |
1059 | recompile some other libraries. The problem is that libio had to be | |
1060 | changed and therefore libraries that are based or depend on the libio | |
1061 | of glibc, e.g. ncurses or slang, need to be recompiled. If you | |
1062 | experience strange segmentation faults in your programs linked against | |
1063 | glibc 2.1, you might need to recompile your libraries. | |
1064 | ||
1065 | Another problem is that older binaries that were linked statically against | |
1066 | glibc 2.0 will reference the older nss modules (libnss_files.so.1 instead of | |
1067 | libnss_files.so.2), so don't remove them. Also, the old glibc-2.0 compiled | |
1068 | static libraries (libfoo.a) which happen to depend on the older libio | |
1069 | behavior will be broken by the glibc 2.1 upgrade. We plan to produce a | |
1070 | compatibility library that people will be able to link in if they want | |
1071 | to compile a static library generated against glibc 2.0 into a program | |
1072 | on a glibc 2.1 system. You just add -lcompat and you should be fine. | |
1073 | ||
1074 | The glibc-compat add-on will provide the libcompat.a library, the older | |
1075 | nss modules, and a few other files. Together, they should make it | |
1076 | possible to do development with old static libraries on a glibc 2.1 | |
8d8c6efa | 1077 | system. This add-on is still in development. You can get it from |
66f6a52b | 1078 | <ftp://alpha.gnu.org/gnu/glibc-compat-2.1.tar.gz> |
b710a6e2 UD |
1079 | but please keep in mind that it is experimental. |
1080 | ||
b7398be5 UD |
1081 | |
1082 | 2.28. Why is extracting files via tar so slow? | |
1083 | ||
1084 | {AJ} Extracting of tar archives might be quite slow since tar has to look up | |
1085 | userid and groupids and doesn't cache negative results. If you have nis or | |
1086 | nisplus in your /etc/nsswitch.conf for the passwd and/or group database, | |
1087 | each file extractions needs a network connection. There are two possible | |
1088 | solutions: | |
1089 | ||
1090 | - do you really need NIS/NIS+ (some Linux distributions add by default | |
1091 | nis/nisplus even if it's not needed)? If not, just remove the entries. | |
1092 | ||
1093 | - if you need NIS/NIS+, use the Name Service Cache Daemon nscd that comes | |
1094 | with glibc 2.1. | |
1095 | ||
2ee511d9 UD |
1096 | |
1097 | 2.29. Compiling programs I get parse errors in libio.h (e.g. "parse error | |
1098 | before `_IO_seekoff'"). How should I fix this? | |
1099 | ||
1100 | {AJ} You might get the following errors when upgrading to glibc 2.1: | |
1101 | ||
1102 | In file included from /usr/include/stdio.h:57, | |
1103 | from ... | |
1104 | /usr/include/libio.h:335: parse error before `_IO_seekoff' | |
1105 | /usr/include/libio.h:335: parse error before `_G_off64_t' | |
1106 | /usr/include/libio.h:336: parse error before `_IO_seekpos' | |
1107 | /usr/include/libio.h:336: parse error before `_G_fpos64_t' | |
1108 | ||
1109 | The problem is a wrong _G_config.h file in your include path. The | |
1110 | _G_config.h file that comes with glibc 2.1 should be used and not one from | |
1111 | libc5 or from a compiler directory. To check which _G_config.h file the | |
1112 | compiler uses, compile your program with `gcc -E ...|grep G_config.h' and | |
1113 | remove that file. Your compiler should pick up the file that has been | |
1114 | installed by glibc 2.1 in your include directory. | |
1115 | ||
4f7ea427 UD |
1116 | |
1117 | 2.30. After upgrading to glibc 2.1, libraries that were compiled against | |
1118 | glibc 2.0.x don't work anymore. | |
1119 | ||
1120 | {AJ} See question 2.27. | |
1121 | ||
61952351 UD |
1122 | \f |
1123 | . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . | |
a5a0310d | 1124 | |
61952351 | 1125 | 3. Source and binary incompatibilities, and what to do about them |
a5a0310d | 1126 | |
61952351 UD |
1127 | 3.1. I expect GNU libc to be 100% source code compatible with |
1128 | the old Linux based GNU libc. Why isn't it like this? | |
a5a0310d | 1129 | |
f12944ec UD |
1130 | {DMT,UD} Not every extension in Linux libc's history was well thought-out. |
1131 | In fact it had a lot of problems with standards compliance and with | |
1132 | cleanliness. With the introduction of a new version number these errors can | |
1133 | now be corrected. Here is a list of the known source code | |
61952351 | 1134 | incompatibilities: |
af6f3906 | 1135 | |
61952351 UD |
1136 | * _GNU_SOURCE: glibc does not make the GNU extensions available |
1137 | automatically. If a program depends on GNU extensions or some | |
1138 | other non-standard functionality, it is necessary to compile it | |
1139 | with the C compiler option -D_GNU_SOURCE, or better, to put | |
1140 | `#define _GNU_SOURCE' at the beginning of your source files, before | |
1141 | any C library header files are included. This difference normally | |
1142 | manifests itself in the form of missing prototypes and/or data type | |
1143 | definitions. Thus, if you get such errors, the first thing you | |
1144 | should do is try defining _GNU_SOURCE and see if that makes the | |
1145 | problem go away. | |
af6f3906 | 1146 | |
61952351 UD |
1147 | For more information consult the file `NOTES' in the GNU C library |
1148 | sources. | |
af6f3906 | 1149 | |
61952351 UD |
1150 | * reboot(): GNU libc sanitizes the interface of reboot() to be more |
1151 | compatible with the interface used on other OSes. reboot() as | |
1152 | implemented in glibc takes just one argument. This argument | |
1153 | corresponds to the third argument of the Linux reboot system call. | |
1154 | That is, a call of the form reboot(a, b, c) needs to be changed into | |
1155 | reboot(c). Beside this the header <sys/reboot.h> defines the needed | |
1156 | constants for the argument. These RB_* constants should be used | |
1157 | instead of the cryptic magic numbers. | |
1158 | ||
1159 | * swapon(): the interface of this function didn't change, but the | |
1160 | prototype is in a separate header file <sys/swap.h>. This header | |
1161 | file also provides the SWAP_* constants defined by <linux/swap.h>; | |
1162 | you should use them for the second argument to swapon(). | |
1163 | ||
1164 | * errno: If a program uses the variable "errno", then it _must_ | |
1165 | include <errno.h>. The old libc often (erroneously) declared this | |
1166 | variable implicitly as a side-effect of including other libc header | |
1167 | files. glibc is careful to avoid such namespace pollution, which, | |
1168 | in turn, means that you really need to include the header files that | |
1169 | you depend on. This difference normally manifests itself in the | |
1170 | form of the compiler complaining about references to an undeclared | |
1171 | symbol "errno". | |
dd7d45e8 | 1172 | |
61952351 UD |
1173 | * Linux-specific syscalls: All Linux system calls now have appropriate |
1174 | library wrappers and corresponding declarations in various header files. | |
1175 | This is because the syscall() macro that was traditionally used to | |
1176 | work around missing syscall wrappers are inherently non-portable and | |
1177 | error-prone. The following table lists all the new syscall stubs, | |
1178 | the header-file declaring their interface and the system call name. | |
dd7d45e8 | 1179 | |
61952351 UD |
1180 | syscall name: wrapper name: declaring header file: |
1181 | ------------- ------------- ---------------------- | |
1182 | bdflush bdflush <sys/kdaemon.h> | |
1183 | syslog ksyslog_ctl <sys/klog.h> | |
dd7d45e8 | 1184 | |
61952351 UD |
1185 | * lpd: Older versions of lpd depend on a routine called _validuser(). |
1186 | The library does not provide this function, but instead provides | |
1187 | __ivaliduser() which has a slightly different interface. Simply | |
1188 | upgrading to a newer lpd should fix this problem (e.g., the 4.4BSD | |
1189 | lpd is known to be working). | |
dd7d45e8 | 1190 | |
61952351 UD |
1191 | * resolver functions/BIND: like on many other systems the functions of |
1192 | the resolver library are not included in libc itself. There is a | |
1193 | separate library libresolv. If you get undefined symbol errors for | |
1194 | symbols starting with `res_*' simply add -lresolv to your linker | |
1195 | command line. | |
dd7d45e8 | 1196 | |
61952351 UD |
1197 | * the `signal' function's behavior corresponds to the BSD semantic and |
1198 | not the SysV semantic as it was in libc-5. The interface on all GNU | |
1199 | systems shall be the same and BSD is the semantic of choice. To use | |
1200 | the SysV behavior simply use `sysv_signal', or define _XOPEN_SOURCE. | |
1201 | See question 3.7 for details. | |
1cab5444 | 1202 | |
1cab5444 | 1203 | |
61952351 UD |
1204 | 3.2. Why does getlogin() always return NULL on my Linux box? |
1205 | ||
f12944ec UD |
1206 | {UD} The GNU C library has a format for the UTMP and WTMP file which differs |
1207 | from what your system currently has. It was extended to fulfill the needs | |
1208 | of the next years when IPv6 is introduced. The record size is different and | |
1209 | some fields have different positions. The files written by functions from | |
1210 | the one library cannot be read by functions from the other library. Sorry, | |
1211 | but this is what a major release is for. It's better to have a cut now than | |
1212 | having no means to support the new techniques later. | |
1cab5444 | 1213 | |
f12944ec UD |
1214 | {MK} There is however a (partial) solution for this problem. Please take a |
1215 | look at the file `login/README.utmpd'. | |
1cab5444 | 1216 | |
6973fc01 | 1217 | |
61952351 UD |
1218 | 3.3. Where are the DST_* constants found in <sys/time.h> on many |
1219 | systems? | |
6973fc01 | 1220 | |
f12944ec UD |
1221 | {UD} These constants come from the old BSD days and are not used anymore |
1222 | (libc5 does not actually implement the handling although the constants are | |
1223 | defined). | |
6973fc01 | 1224 | |
f12944ec | 1225 | Instead GNU libc contains zone database support and compatibility code for |
8b4a4715 | 1226 | POSIX TZ environment variable handling. For former is very much preferred |
7d1de115 | 1227 | (see question 4.3). |
6973fc01 UD |
1228 | |
1229 | ||
61952351 UD |
1230 | 3.4. The prototypes for `connect', `accept', `getsockopt', |
1231 | `setsockopt', `getsockname', `getpeername', `send', | |
1232 | `sendto', and `recvfrom' are different in GNU libc from | |
1233 | any other system I saw. This is a bug, isn't it? | |
f4017d20 | 1234 | |
f12944ec UD |
1235 | {UD} No, this is no bug. This version of GNU libc already follows the new |
1236 | Single Unix specifications (and I think the POSIX.1g draft which adopted the | |
1237 | solution). The type for a parameter describing a size is now `socklen_t', a | |
1238 | new type. | |
f4017d20 | 1239 | |
f4017d20 | 1240 | |
61952351 UD |
1241 | 3.5. On Linux I've got problems with the declarations in Linux |
1242 | kernel headers. | |
f4017d20 | 1243 | |
f12944ec UD |
1244 | {UD,AJ} On Linux, the use of kernel headers is reduced to the minimum. This |
1245 | gives Linus the ability to change the headers more freely. Also, user | |
8f1c9b09 | 1246 | programs are now insulated from changes in the size of kernel data |
f12944ec | 1247 | structures. |
f4017d20 | 1248 | |
f12944ec UD |
1249 | For example, the sigset_t type is 32 or 64 bits wide in the kernel. In |
1250 | glibc it is 1024 bits wide. This guarantees that when the kernel gets a | |
1251 | bigger sigset_t (for POSIX.1e realtime support, say) user programs will not | |
1252 | have to be recompiled. Consult the header files for more information about | |
1253 | the changes. | |
61952351 | 1254 | |
f12944ec UD |
1255 | Therefore you shouldn't include Linux kernel header files directly if glibc |
1256 | has defined a replacement. Otherwise you might get undefined results because | |
1257 | of type conflicts. | |
f4017d20 | 1258 | |
f4017d20 | 1259 | |
61952351 UD |
1260 | 3.6. I don't include any kernel headers myself but the compiler |
1261 | still complains about redeclarations of types in the kernel | |
1262 | headers. | |
1263 | ||
f12944ec UD |
1264 | {UD} The kernel headers before Linux 2.1.61 and 2.0.32 don't work correctly |
1265 | with glibc. Compiling C programs is possible in most cases but C++ programs | |
1266 | have (due to the change of the name lookups for `struct's) problems. One | |
1267 | prominent example is `struct fd_set'. | |
61952351 | 1268 | |
f12944ec UD |
1269 | There might be some problems left but 2.1.61/2.0.32 fix most of the known |
1270 | ones. See the BUGS file for other known problems. | |
61952351 UD |
1271 | |
1272 | ||
1273 | 3.7. Why don't signals interrupt system calls anymore? | |
1274 | ||
f12944ec UD |
1275 | {ZW} By default GNU libc uses the BSD semantics for signal(), unlike Linux |
1276 | libc 5 which used System V semantics. This is partially for compatibility | |
1277 | with other systems and partially because the BSD semantics tend to make | |
1278 | programming with signals easier. | |
f4017d20 UD |
1279 | |
1280 | There are three differences: | |
1281 | ||
1282 | * BSD-style signals that occur in the middle of a system call do not | |
1283 | affect the system call; System V signals cause the system call to | |
1284 | fail and set errno to EINTR. | |
1285 | ||
1286 | * BSD signal handlers remain installed once triggered. System V signal | |
1287 | handlers work only once, so one must reinstall them each time. | |
1288 | ||
1289 | * A BSD signal is blocked during the execution of its handler. In other | |
1290 | words, a handler for SIGCHLD (for example) does not need to worry about | |
61952351 | 1291 | being interrupted by another SIGCHLD. It may, however, be interrupted |
f4017d20 UD |
1292 | by other signals. |
1293 | ||
1294 | There is general consensus that for `casual' programming with signals, the | |
1295 | BSD semantics are preferable. You don't need to worry about system calls | |
1296 | returning EINTR, and you don't need to worry about the race conditions | |
1297 | associated with one-shot signal handlers. | |
1298 | ||
1299 | If you are porting an old program that relies on the old semantics, you can | |
1300 | quickly fix the problem by changing signal() to sysv_signal() throughout. | |
1301 | Alternatively, define _XOPEN_SOURCE before including <signal.h>. | |
1302 | ||
1303 | For new programs, the sigaction() function allows you to specify precisely | |
1304 | how you want your signals to behave. All three differences listed above are | |
1305 | individually switchable on a per-signal basis with this function. | |
1306 | ||
f12944ec UD |
1307 | If all you want is for one specific signal to cause system calls to fail and |
1308 | return EINTR (for example, to implement a timeout) you can do this with | |
f4017d20 UD |
1309 | siginterrupt(). |
1310 | ||
1311 | ||
61952351 UD |
1312 | 3.8. I've got errors compiling code that uses certain string |
1313 | functions. Why? | |
1314 | ||
f12944ec | 1315 | {AJ} glibc 2.1 has special string functions that are faster than the normal |
fdacb17d | 1316 | library functions. Some of the functions are additionally implemented as |
a25f2023 UD |
1317 | inline functions and others as macros. This might lead to problems with |
1318 | existing codes but it is explicitly allowed by ISO C. | |
04be94a8 | 1319 | |
04be94a8 | 1320 | The optimized string functions are only used when compiling with |
fdacb17d | 1321 | optimizations (-O1 or higher). The behavior can be changed with two feature |
f12944ec | 1322 | macros: |
61952351 UD |
1323 | |
1324 | * __NO_STRING_INLINES: Don't do any string optimizations. | |
1325 | * __USE_STRING_INLINES: Use assembly language inline functions (might | |
1326 | increase code size dramatically). | |
04be94a8 | 1327 | |
f12944ec UD |
1328 | Since some of these string functions are now additionally defined as macros, |
1329 | code like "char *strncpy();" doesn't work anymore (and is unnecessary, since | |
fdacb17d | 1330 | <string.h> has the necessary declarations). Either change your code or |
f12944ec | 1331 | define __NO_STRING_INLINES. |
04be94a8 | 1332 | |
f12944ec UD |
1333 | {UD} Another problem in this area is that gcc still has problems on machines |
1334 | with very few registers (e.g., ix86). The inline assembler code can require | |
1335 | almost all the registers and the register allocator cannot always handle | |
1336 | this situation. | |
04be94a8 | 1337 | |
61952351 | 1338 | One can disable the string optimizations selectively. Instead of writing |
04be94a8 UD |
1339 | |
1340 | cp = strcpy (foo, "lkj"); | |
1341 | ||
1342 | one can write | |
1343 | ||
1344 | cp = (strcpy) (foo, "lkj"); | |
1345 | ||
61952351 UD |
1346 | This disables the optimization for that specific call. |
1347 | ||
4775243a UD |
1348 | |
1349 | 3.9. I get compiler messages "Initializer element not constant" with | |
1350 | stdin/stdout/stderr. Why? | |
1351 | ||
1352 | {RM,AJ} Constructs like: | |
66f6a52b | 1353 | static FILE *InPtr = stdin; |
4775243a | 1354 | |
fdacb17d UD |
1355 | lead to this message. This is correct behaviour with glibc since stdin is |
1356 | not a constant expression. Please note that a strict reading of ISO C does | |
f12944ec | 1357 | not allow above constructs. |
4775243a | 1358 | |
f12944ec UD |
1359 | One of the advantages of this is that you can assign to stdin, stdout, and |
1360 | stderr just like any other global variable (e.g. `stdout = my_stream;'), | |
1361 | which can be very useful with custom streams that you can write with libio | |
fdacb17d | 1362 | (but beware this is not necessarily portable). The reason to implement it |
f12944ec | 1363 | this way were versioning problems with the size of the FILE structure. |
4775243a | 1364 | |
fdacb17d UD |
1365 | To fix those programs you've got to initialize the variable at run time. |
1366 | This can be done, e.g. in main, like: | |
1367 | ||
66f6a52b UD |
1368 | static FILE *InPtr; |
1369 | int main(void) | |
1370 | { | |
1371 | InPtr = stdin; | |
1372 | } | |
fdacb17d UD |
1373 | |
1374 | or by constructors (beware this is gcc specific): | |
1375 | ||
66f6a52b UD |
1376 | static FILE *InPtr; |
1377 | static void inPtr_construct (void) __attribute__((constructor)); | |
1378 | static void inPtr_construct (void) { InPtr = stdin; } | |
fdacb17d | 1379 | |
4775243a UD |
1380 | |
1381 | 3.10. I can't compile with gcc -traditional (or | |
1382 | -traditional-cpp). Why? | |
1383 | ||
1384 | {AJ} glibc2 does break -traditional and -traditonal-cpp - and will continue | |
fdacb17d | 1385 | to do so. For example constructs of the form: |
f12944ec | 1386 | |
66f6a52b UD |
1387 | enum {foo |
1388 | #define foo foo | |
1389 | } | |
f12944ec UD |
1390 | |
1391 | are useful for debugging purposes (you can use foo with your debugger that's | |
1392 | why we need the enum) and for compatibility (other systems use defines and | |
1393 | check with #ifdef). | |
4775243a UD |
1394 | |
1395 | ||
1396 | 3.11. I get some errors with `gcc -ansi'. Isn't glibc ANSI compatible? | |
1397 | ||
1398 | {AJ} The GNU C library is compatible with the ANSI/ISO C standard. If | |
f12944ec | 1399 | you're using `gcc -ansi', the glibc includes which are specified in the |
fdacb17d | 1400 | standard follow the standard. The ANSI/ISO C standard defines what has to be |
f12944ec UD |
1401 | in the include files - and also states that nothing else should be in the |
1402 | include files (btw. you can still enable additional standards with feature | |
1403 | flags). | |
4775243a | 1404 | |
f12944ec UD |
1405 | The GNU C library is conforming to ANSI/ISO C - if and only if you're only |
1406 | using the headers and library functions defined in the standard. | |
4775243a | 1407 | |
a35cb74d UD |
1408 | |
1409 | 3.12. I can't access some functions anymore. nm shows that they do | |
1410 | exist but linking fails nevertheless. | |
1411 | ||
f12944ec UD |
1412 | {AJ} With the introduction of versioning in glibc 2.1 it is possible to |
1413 | export only those identifiers (functions, variables) that are really needed | |
1414 | by application programs and by other parts of glibc. This way a lot of | |
1415 | internal interfaces are now hidden. nm will still show those identifiers | |
1416 | but marking them as internal. ISO C states that identifiers beginning with | |
1417 | an underscore are internal to the libc. An application program normally | |
1418 | shouldn't use those internal interfaces (there are exceptions, | |
1419 | e.g. __ivaliduser). If a program uses these interfaces, it's broken. These | |
1420 | internal interfaces might change between glibc releases or dropped | |
1421 | completely. | |
a35cb74d | 1422 | |
a5f4e34a UD |
1423 | |
1424 | 3.13. When using the db-2 library which comes with glibc is used in | |
1425 | the Perl db modules the testsuite is not passed. This did not | |
1426 | happen with db-1, gdbm, or ndbm. | |
1427 | ||
3877d9ea UD |
1428 | {MK} Db-2 does not support zero-sized keys. The Perl testsuite |
1429 | tests the support for zero-sized keys and therefore fails when db-2 is | |
1430 | used. The Perl folks are looking for a solution, but thus far have | |
1431 | not found a satisfactory one. | |
a5f4e34a | 1432 | |
5148d49f UD |
1433 | |
1434 | 3.14. The pow() inline function I get when including <math.h> is broken. | |
1435 | I get segmentation faults when I run the program. | |
1436 | ||
1437 | {UD} Nope, the implementation is correct. The problem is with egcs version | |
1438 | prior to 1.1. I.e., egcs 1.0 to 1.0.3 are all broken (at least on Intel). | |
1439 | If you have to use this compiler you must define __NO_MATH_INLINES before | |
1440 | including <math.h> to prevent the inline functions from being used. egcs 1.1 | |
1441 | fixes the problem. I don't know about gcc 2.8 and 2.8.1. | |
1442 | ||
05f732b3 UD |
1443 | |
1444 | 3.15. The sys/sem.h file lacks the definition of `union semun'. | |
1445 | ||
1446 | {UD} Nope. This union has to be provided by the user program. Former glibc | |
1447 | versions defined this but it was an error since it does not make much sense | |
1448 | when thinking about it. The standards describing the System V IPC functions | |
1449 | define it this way and therefore programs must be adopted. | |
1450 | ||
33127459 UD |
1451 | |
1452 | 3.16. Why has <netinet/ip_fw.h> disappeared? | |
1453 | ||
1454 | {AJ} The corresponding Linux kernel data structures and constants are | |
b710a6e2 | 1455 | totally different in Linux 2.0 and Linux 2.2. This situation has to be |
33127459 UD |
1456 | taken care in user programs using the firewall structures and therefore |
1457 | those programs (ipfw is AFAIK the only one) should deal with this problem | |
1458 | themselves. | |
1459 | ||
28ab8526 UD |
1460 | |
1461 | 3.17. I get floods of warnings when I use -Wconversion and include | |
1462 | <string.h> or <math.h>. | |
1463 | ||
1464 | {ZW} <string.h> and <math.h> intentionally use prototypes to override | |
1465 | argument promotion. -Wconversion warns about all these. You can safely | |
1466 | ignore the warnings. | |
1467 | ||
1468 | -Wconversion isn't really intended for production use, only for shakedown | |
1469 | compiles after converting an old program to standard C. | |
1470 | ||
5ff1a70a UD |
1471 | |
1472 | 3.18. After upgrading to glibc 2.1, I receive errors about | |
1473 | unresolved symbols, like `_dl_initial_searchlist' and can not | |
1474 | execute any binaries. What went wrong? | |
1475 | ||
1476 | {AJ} This normally happens if your libc and ld (dynamic linker) are from | |
1477 | different releases of glibc. For example, the dynamic linker | |
1478 | /lib/ld-linux.so.2 comes from glibc 2.0.x, but the version of libc.so.6 is | |
1479 | from glibc 2.1. | |
1480 | ||
1481 | The path /lib/ld-linux.so.2 is hardcoded in every glibc2 binary but | |
1482 | libc.so.6 is searched via /etc/ld.so.cache and in some special directories | |
1483 | like /lib and /usr/lib. If you run configure with another prefix than /usr | |
1484 | and put this prefix before /lib in /etc/ld.so.conf, your system will break. | |
1485 | ||
1486 | So what can you do? Either of the following should work: | |
1487 | ||
1488 | * Run `configure' with the same prefix argument you've used for glibc 2.0.x | |
1489 | so that the same paths are used. | |
1490 | * Replace /lib/ld-linux.so.2 with a link to the dynamic linker from glibc | |
1491 | 2.1. | |
1492 | ||
1493 | You can even call the dynamic linker by hand if everything fails. You've | |
1494 | got to set LD_LIBRARY_PATH so that the corresponding libc is found and also | |
1495 | need to provide an absolute path to your binary: | |
1496 | ||
1497 | LD_LIBRARY_PATH=<path-where-libc.so.6-lives> \ | |
1498 | <path-where-corresponding-dynamic-linker-lives>/ld-linux.so.2 \ | |
1499 | <path-to-binary>/binary | |
1500 | ||
1501 | For example `LD_LIBRARY_PATH=/libold /libold/ld-linux.so.2 /bin/mv ...' | |
1502 | might be useful in fixing a broken system (if /libold contains dynamic | |
1503 | linker and corresponding libc). | |
1504 | ||
1505 | With that command line no path is used. To further debug problems with the | |
1506 | dynamic linker, use the LD_DEBUG environment variable, e.g. | |
1507 | `LD_DEBUG=help echo' for the help text. | |
1508 | ||
1509 | If you just want to test this release, don't put the lib directory in | |
1510 | /etc/ld.so.conf. You can call programs directly with full paths (as above). | |
1511 | When compiling new programs against glibc 2.1, you've got to specify the | |
1512 | correct paths to the compiler (option -I with gcc) and linker (options | |
1513 | --dynamic-linker, -L and --rpath). | |
1514 | ||
61952351 | 1515 | |
7d1de115 UD |
1516 | 3.19. bonnie reports that char i/o with glibc 2 is much slower than with |
1517 | libc5. What can be done? | |
1518 | ||
1519 | {AJ} The GNU C library uses thread safe functions by default and libc5 used | |
1520 | non thread safe versions. The non thread safe functions have in glibc the | |
1521 | suffix `_unlocked', for details check <stdio.h>. Using `putc_unlocked' etc. | |
1522 | instead of `putc' should give nearly the same speed with bonnie (bonnie is a | |
1523 | benchmark program for measuring disk access). | |
9f6b6d8d | 1524 | |
b93492aa UD |
1525 | |
1526 | 3.20. Programs compiled with glibc 2.1 can't read db files made with glibc | |
1527 | 2.0. What has changed that programs like rpm break? | |
1528 | ||
1529 | {AJ} The GNU C library 2.1 uses db2 instead of db1 which was used in version | |
1530 | 2.0. The internal formats of the actual db files are different. To convert | |
1531 | the db files from db1 format to db2 format, you can use the programs | |
1532 | `db_dump185' and `db_load'. Alternativly programs can be linked with db1 | |
1533 | using `-ldb1' instead of linking with db2 which uses `-ldb'. Linking with | |
1534 | db1 might be preferable if older programs need to access the db file. | |
1535 | ||
1536 | db2 supports the old db1 programming interface and also a new programming | |
1537 | interface. For compilation with the old API, <db_185.h> has to be included | |
1538 | (and not <db.h>) and you can link with either `-ldb1' or `-ldb' for either | |
1539 | of the db formats. | |
1540 | ||
b5a9efcd UD |
1541 | |
1542 | 3.21. Autoconf's AC_CHECK_FUNC macro reports that a function exists, but | |
1543 | when I try to use it, it always returns -1 and sets errno to ENOSYS. | |
1544 | ||
1545 | {ZW} You are using a 2.0 Linux kernel, and the function you are trying to | |
1546 | use is only implemented in 2.1/2.2. Libc considers this to be a function | |
1547 | which exists, because if you upgrade to a 2.2 kernel, it will work. One | |
1548 | such function is sigaltstack. | |
1549 | ||
1550 | Your program should check at runtime whether the function works, and | |
1551 | implement a fallback. Note that Autoconf cannot detect unimplemented | |
1552 | functions in other systems' C libraries, so you need to do this anyway. | |
1553 | ||
1554 | ||
1555 | 3.22. My program segfaults when I call fclose() on the FILE* returned | |
1556 | from setmntent(). Is this a glibc bug? | |
1557 | ||
1558 | {GK} No. Don't do this. Use endmntent(), that's what it's for. | |
1559 | ||
1560 | In general, you should use the correct deallocation routine. For instance, | |
1561 | if you open a file using fopen(), you should deallocate the FILE * using | |
1562 | fclose(), not free(), even though the FILE * is also a pointer. | |
1563 | ||
1564 | In the case of setmntent(), it may appear to work in most cases, but it | |
1565 | won't always work. Unfortunately, for compatibility reasons, we can't | |
1566 | change the return type of setmntent() to something other than FILE *. | |
1567 | ||
9f6b6d8d UD |
1568 | \f |
1569 | . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . | |
1570 | ||
7d1de115 | 1571 | 4. Miscellaneous |
61952351 | 1572 | |
7d1de115 | 1573 | 4.1. After I changed configure.in I get `Autoconf version X.Y. |
61952351 UD |
1574 | or higher is required for this script'. What can I do? |
1575 | ||
1576 | {UD} You have to get the specified autoconf version (or a later one) | |
a35cb74d | 1577 | from your favorite mirror of ftp.gnu.org. |
61952351 | 1578 | |
04be94a8 | 1579 | |
7d1de115 | 1580 | 4.2. When I try to compile code which uses IPv6 headers and |
61952351 UD |
1581 | definitions on my Linux 2.x.y system I am in trouble. |
1582 | Nothing seems to work. | |
1583 | ||
f12944ec UD |
1584 | {UD} The problem is that IPv6 development still has not reached a point |
1585 | where the headers are stable. There are still lots of incompatible changes | |
1586 | made and the libc headers have to follow. | |
61952351 | 1587 | |
348ed515 UD |
1588 | {PB} The 2.1 release of GNU libc aims to comply with the current versions of |
1589 | all the relevant standards. The IPv6 support libraries for older Linux | |
1590 | systems used a different naming convention and so code written to work with | |
1591 | them may need to be modified. If the standards make incompatible changes in | |
1592 | the future then the libc may need to change again. | |
1593 | ||
1594 | IPv6 will not work with a 2.0.x kernel. When kernel 2.2 is released it | |
1595 | should contain all the necessary support; until then you should use the | |
1596 | latest 2.1.x release you can find. As of 98/11/26 the currently recommended | |
1597 | kernel for IPv6 is 2.1.129. | |
1598 | ||
1599 | Also, as of the 2.1 release the IPv6 API provided by GNU libc is not | |
1600 | 100% complete. In particular the getipnodebyname and getipnodebyaddr | |
1601 | functions are not implemented. | |
04be94a8 | 1602 | |
ff44f2a5 | 1603 | |
7d1de115 | 1604 | 4.3. When I set the timezone by setting the TZ environment variable |
ff44f2a5 UD |
1605 | to EST5EDT things go wrong since glibc computes the wrong time |
1606 | from this information. | |
1607 | ||
f12944ec UD |
1608 | {UD} The problem is that people still use the braindamaged POSIX method to |
1609 | select the timezone using the TZ environment variable with a format EST5EDT | |
8b4a4715 UD |
1610 | or whatever. People, if you insist on using TZ instead of the timezone |
1611 | database (see below), read the POSIX standard, the implemented behaviour is | |
f12944ec UD |
1612 | correct! What you see is in fact the result of the decisions made while |
1613 | POSIX.1 was created. We've only implemented the handling of TZ this way to | |
1614 | be POSIX compliant. It is not really meant to be used. | |
1615 | ||
1616 | The alternative approach to handle timezones which is implemented is the | |
1617 | correct one to use: use the timezone database. This avoids all the problems | |
1618 | the POSIX method has plus it is much easier to use. Simply run the tzselect | |
1619 | shell script, answer the question and use the name printed in the end by | |
8b4a4715 UD |
1620 | making a symlink /etc/localtime pointing to /usr/share/zoneinfo/NAME (NAME |
1621 | is the returned value from tzselect). That's all. You never again have to | |
1622 | worry. | |
f12944ec UD |
1623 | |
1624 | So, please avoid sending bug reports about time related problems if you use | |
1625 | the POSIX method and you have not verified something is really broken by | |
1626 | reading the POSIX standards. | |
ff44f2a5 | 1627 | |
fdacb17d | 1628 | |
7d1de115 | 1629 | 4.4. What other sources of documentation about glibc are available? |
fdacb17d UD |
1630 | |
1631 | {AJ} The FSF has a page about the GNU C library at | |
1632 | <http://www.gnu.org/software/libc/>. The problem data base of open and | |
1633 | solved bugs in GNU libc is available at | |
1634 | <http://www-gnats.gnu.org:8080/cgi-bin/wwwgnats.pl>. Eric Green has written | |
1635 | a HowTo for converting from Linux libc5 to glibc2. The HowTo is accessable | |
1636 | via the FSF page and at <http://www.imaxx.net/~thrytis/glibc>. Frodo | |
1637 | Looijaard describes a different way installing glibc2 as secondary libc at | |
1638 | <http://huizen.dds.nl/~frodol/glibc>. | |
1639 | ||
1640 | Please note that this is not a complete list. | |
1641 | ||
348ed515 | 1642 | |
7d1de115 | 1643 | 4.5. The timezone string for Sydney/Australia is wrong since even when |
348ed515 UD |
1644 | daylight saving time is in effect the timezone string is EST. |
1645 | ||
1646 | {UD} The problem for some timezones is that the local authorities decided | |
1647 | to use the term "summer time" instead of "daylight saving time". In this | |
1648 | case the abbreviation character `S' is the same as the standard one. So, | |
1649 | for Sydney we have | |
1650 | ||
1651 | Eastern Standard Time = EST | |
1652 | Eastern Summer Time = EST | |
1653 | ||
1654 | Great! To get this bug fixed convince the authorities to change the laws | |
1655 | and regulations of the country this effects. glibc behaves correctly. | |
1656 | ||
eeabe877 | 1657 | |
7d1de115 | 1658 | 4.6. I've build make 3.77 against glibc 2.1 and now make gets |
eeabe877 UD |
1659 | segmentation faults. |
1660 | ||
1661 | {AJ} GNU make 3.77 has support for 64 bit filesystems which is slightly | |
1662 | broken (and one of the new features in the GNU C library 2.1 is 64 bit | |
1663 | filesystem support :-( ). To get a working make you can use either make | |
1664 | 3.75 or patch 3.77. A working patch is available via RedHat's Rawhide server | |
1665 | (ftp://rawhide.redhat.com/SRPMS/SRPMS/make-3.77-*src.rpm). | |
1666 | ||
f8cac037 | 1667 | \f |
61952351 UD |
1668 | ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ |
1669 | ||
f8cac037 RM |
1670 | Answers were given by: |
1671 | {UD} Ulrich Drepper, <drepper@cygnus.com> | |
613a76ff | 1672 | {DMT} David Mosberger-Tang, <davidm@AZStarNet.com> |
dd7d45e8 | 1673 | {RM} Roland McGrath, <roland@gnu.org> |
1f205a47 | 1674 | {AJ} Andreas Jaeger, <aj@arthur.rhein-neckar.de> |
22d57dd3 | 1675 | {EY} Eric Youngdale, <eric@andante.jic.com> |
a5a0310d | 1676 | {PB} Phil Blundell, <Philip.Blundell@pobox.com> |
af6f3906 | 1677 | {MK} Mark Kettenis, <kettenis@phys.uva.nl> |
f4017d20 | 1678 | {ZW} Zack Weinberg, <zack@rabi.phys.columbia.edu> |
612fdf25 | 1679 | {TK} Thorsten Kukuk, <kukuk@suse.de> |
8619129f | 1680 | {GK} Geoffrey Keating, <geoffk@ozemail.com.au> |
a35cb74d | 1681 | {HJ} H.J. Lu, <hjl@gnu.org> |
b710a6e2 | 1682 | {CG} Cristian Gafton, <gafton@redhat.com> |
f8cac037 RM |
1683 | \f |
1684 | Local Variables: | |
61952351 UD |
1685 | mode:outline |
1686 | outline-regexp:"\\?" | |
f12944ec | 1687 | fill-column:76 |
f8cac037 | 1688 | End: |