08b6c5deea
fixed handling of dead fastcgi processes if deamonized
Jan Kneschke2005-03-02 13:49:08 +00:00
9a5cf4a2ed
fixed handling of dead fastcgi processes if deamonized
Jan Kneschke2005-03-02 13:49:08 +00:00
5bb78266fa
don't touch the exit-conditions for the loop directly (reverting a this part of one of the last changesets)
Jan Kneschke2005-03-02 11:31:21 +00:00
853b4f5b06
don't touch the exit-conditions for the loop directly (reverting a this part of one of the last changesets)
Jan Kneschke2005-03-02 11:31:21 +00:00
40be6f2b40
seperated the tests for each section and let run-tests.pl use Test::Harness
Jan Kneschke2005-03-02 11:27:02 +00:00
40e8c6714a
seperated the tests for each section and let run-tests.pl use Test::Harness
Jan Kneschke2005-03-02 11:27:02 +00:00
2554ab151f
added | to the non-rfc characters in URI
Jan Kneschke2005-03-02 09:11:42 +00:00
3caa67a131
added | to the non-rfc characters in URI
Jan Kneschke2005-03-02 09:11:42 +00:00
3788f1ee2f
updated ChangeLog and NEWS from timeline
Jan Kneschke2005-03-01 23:28:52 +00:00
06c35abe4a
updated ChangeLog and NEWS from timeline
Jan Kneschke2005-03-01 23:28:52 +00:00
9bca2cdf6f
this one belongs to the last changeset
Jan Kneschke2005-02-28 10:54:23 +00:00
26b184ba6f
this one belongs to the last changeset
Jan Kneschke2005-02-28 10:54:23 +00:00
4c814e01fa
* If sizeof(long) == sizeof(off_t), buffer_{append,copy}_off_t() are only macros to buffer_{append,copy}_long() * ltostr() returns the string length instead of always 0 * Don't check return value of buffer_prepare_append(buffer *b), since it only returns -1 if b == NULL, which we do a few lines above anyway. * Improved buffer_path_simplify(). No "dot_stack" required anymore. Operation can also be performed inplace. * Check errno also against EACCES at pidfile-unlink for not logging a "Permission denied".
Jan Kneschke2005-02-28 10:38:16 +00:00
75fae49e37
* If sizeof(long) == sizeof(off_t), buffer_{append,copy}_off_t() are only macros to buffer_{append,copy}_long() * ltostr() returns the string length instead of always 0 * Don't check return value of buffer_prepare_append(buffer *b), since it only returns -1 if b == NULL, which we do a few lines above anyway. * Improved buffer_path_simplify(). No "dot_stack" required anymore. Operation can also be performed inplace. * Check errno also against EACCES at pidfile-unlink for not logging a "Permission denied".
Jan Kneschke2005-02-28 10:38:16 +00:00
390fa3c979
* Changed AM_CONFIG_HEADER to AC_CONFIG_HEADER. The Manual says: Older versions of Automake required the use of AM_CONFIG_HEADER; this is no longer the case today.
Jan Kneschke2005-02-28 10:36:12 +00:00
d9fd24a0f7
* Changed AM_CONFIG_HEADER to AC_CONFIG_HEADER. The Manual says: Older versions of Automake required the use of AM_CONFIG_HEADER; this is no longer the case today.
Jan Kneschke2005-02-28 10:36:12 +00:00
56690eb7a1
tolerate \n\n instead of \r\n\r\n and handle overlapping EOL correctly
Jan Kneschke2005-02-28 09:35:06 +00:00
aa63120540
tolerate \n\n instead of \r\n\r\n and handle overlapping EOL correctly
Jan Kneschke2005-02-28 09:35:06 +00:00
2a343ae169
replace inet_addr() by inet_aton() as the man-page on linux recommends and as solaris likes for SunForte
Jan Kneschke2005-02-22 00:18:04 +00:00
bc7f12ca71
replace inet_addr() by inet_aton() as the man-page on linux recommends and as solaris likes for SunForte
Jan Kneschke2005-02-22 00:18:04 +00:00
3e88e04e9a
fix segfault in log_error_write if host is empty
Jan Kneschke2005-02-22 00:17:11 +00:00
44997909b1
fix segfault in log_error_write if host is empty
Jan Kneschke2005-02-22 00:17:11 +00:00
Jan Kneschke