summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>1999-05-04 12:24:18 +0000
committerUlrich Drepper <drepper@redhat.com>1999-05-04 12:24:18 +0000
commitb5a9efcd9342681200280ef9f764b744d62b62ce (patch)
tree71c18e1855fa1e53ba3e002eb6643c318c27a90b
parent27aa0631c73ee805519f3ac078eaef460f1b4bc3 (diff)
Update.
* elf/dl-load.c (expand_dynamic_string_token): Rewrite to loose st variable.
-rw-r--r--ChangeLog3
-rw-r--r--FAQ39
-rw-r--r--FAQ.in13
-rw-r--r--elf/dl-load.c6
4 files changed, 53 insertions, 8 deletions
diff --git a/ChangeLog b/ChangeLog
index a126fe9a28..4bb25b1394 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,8 @@
1999-05-04 Ulrich Drepper <drepper@cygnus.com>
+ * elf/dl-load.c (expand_dynamic_string_token): Rewrite to loose st
+ variable.
+
* elf/dl-load.c (expand_dynamic_string_token): Recognize { }
around DST. Correctly ignore ORIGIN IN SUID binaries.
diff --git a/FAQ b/FAQ
index 65d1e7d3d7..69edb3d48e 100644
--- a/FAQ
+++ b/FAQ
@@ -147,6 +147,10 @@ please let me know.
libc5. What can be done?
3.20. Programs compiled with glibc 2.1 can't read db files made with glibc
2.0. What has changed that programs like rpm break?
+3.21. Autoconf's AC_CHECK_FUNC macro reports that a function exists, but
+ when I try to use it, it always returns -1 and sets errno to ENOSYS.
+3.22. My program segfaults when I call fclose() on the FILE* returned
+ from setmntent(). Is this a glibc bug?
4. Miscellaneous
@@ -310,10 +314,10 @@ Binutils 2.9.1.0.16 or later is also required.
as much as 400MB).
* plenty of time. Compiling just the shared and static libraries for
- i?86-linux takes approximately 1h on an i586@133, or 2.5h on
- i486@66, or 4.5h on i486@33. Multiply this by 1.5 or 2.0 if you
- build profiling and/or the highly optimized version as well. For
- Hurd systems times are much higher.
+ i?86-linux takes approximately 1h on an AMD-K6@225MHz w/ 96MB of RAM,
+ 45mins on a Celeron@400MHz w/ 128MB, and 55mins on a Alpha@533MHz w/ 256MB.
+ Multiply this by 1.5 or 2.0 if you build profiling and/or the highly
+ optimized version as well. For Hurd systems times are much higher.
You should avoid compiling in a NFS mounted filesystem. This is
very slow.
@@ -1530,6 +1534,33 @@ interface. For compilation with the old API, <db_185.h> has to be included
(and not <db.h>) and you can link with either `-ldb1' or `-ldb' for either
of the db formats.
+
+3.21. Autoconf's AC_CHECK_FUNC macro reports that a function exists, but
+ when I try to use it, it always returns -1 and sets errno to ENOSYS.
+
+{ZW} You are using a 2.0 Linux kernel, and the function you are trying to
+use is only implemented in 2.1/2.2. Libc considers this to be a function
+which exists, because if you upgrade to a 2.2 kernel, it will work. One
+such function is sigaltstack.
+
+Your program should check at runtime whether the function works, and
+implement a fallback. Note that Autoconf cannot detect unimplemented
+functions in other systems' C libraries, so you need to do this anyway.
+
+
+3.22. My program segfaults when I call fclose() on the FILE* returned
+ from setmntent(). Is this a glibc bug?
+
+{GK} No. Don't do this. Use endmntent(), that's what it's for.
+
+In general, you should use the correct deallocation routine. For instance,
+if you open a file using fopen(), you should deallocate the FILE * using
+fclose(), not free(), even though the FILE * is also a pointer.
+
+In the case of setmntent(), it may appear to work in most cases, but it
+won't always work. Unfortunately, for compatibility reasons, we can't
+change the return type of setmntent() to something other than FILE *.
+
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
diff --git a/FAQ.in b/FAQ.in
index 0b950c81c5..6a841a07a3 100644
--- a/FAQ.in
+++ b/FAQ.in
@@ -1326,6 +1326,19 @@ Your program should check at runtime whether the function works, and
implement a fallback. Note that Autoconf cannot detect unimplemented
functions in other systems' C libraries, so you need to do this anyway.
+?? My program segfaults when I call fclose() on the FILE* returned
+ from setmntent(). Is this a glibc bug?
+
+{GK} No. Don't do this. Use endmntent(), that's what it's for.
+
+In general, you should use the correct deallocation routine. For instance,
+if you open a file using fopen(), you should deallocate the FILE * using
+fclose(), not free(), even though the FILE * is also a pointer.
+
+In the case of setmntent(), it may appear to work in most cases, but it
+won't always work. Unfortunately, for compatibility reasons, we can't
+change the return type of setmntent() to something other than FILE *.
+
? Miscellaneous
diff --git a/elf/dl-load.c b/elf/dl-load.c
index bdae4ba1da..46f0b67567 100644
--- a/elf/dl-load.c
+++ b/elf/dl-load.c
@@ -158,13 +158,12 @@ expand_dynamic_string_token (struct link_map *l, const char *s)
resulting string is and then we copy it over. Since this is now
frequently executed operation we are looking here not for performance
but rather for code size. */
- const char *st, *sf;
+ const char *sf;
size_t cnt = 0;
size_t origin_len;
size_t total;
char *result, *last_elem, *wp;
- st = s;
sf = strchr (s, '$');
while (sf != NULL)
{
@@ -182,8 +181,7 @@ expand_dynamic_string_token (struct link_map *l, const char *s)
&& (len = 11) != 0))))
++cnt;
- st = sf + len;
- sf = strchr (st, '$');
+ sf = strchr (sf + len, '$');
}
/* If we do not have to replace anything simply copy the string. */