lib/libhttp
1
0
Fork 0
mirror of https://github.com/lammertb/libhttp.git synced 2025-12-22 04:02:04 +03:00
Code Activity
3,078 Commits 1 Branch 1 Tag
d8d9dd5fff51a51a6c0303a8a093ef4932e3d68b
Commit Graph

3078 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
bel
d8d9dd5fff Read client certificate information (Step 3/3) 2016-07-24 22:38:05 +02:00
bel
aeb8721503 Adapt test to changes in remove_double_dots_and_double_slashes 2016-07-23 22:18:29 +02:00
bel
a84aa1b5ac Read client certificate information (Step 2/3) 2016-07-23 00:31:47 +02:00
bel
5f13e9894b Skip leading .. in remove_double_dots_and_double_slashes 
See #327. There was not really a directory disclosure before, since a request to .. will fail already in process_new_connection with error "Invalid URI: [..]".
 2016-07-22 22:53:01 +02:00
bel
0a2baa7de1 Fix mg_calloc_ex if MEMORY_DEBUGGING enabled (#328) 2016-07-22 22:25:13 +02:00
bel
ee488b231b Read client certificate information (Step 1/3) 2016-07-21 00:07:20 +02:00
bel
a7f49cc7fd Use OS independent atomic function instead of Windows specific 2016-07-15 20:26:26 +02:00
bel
c262166b3f Use atomic operation for locking variable and format code 2016-07-14 20:35:26 +02:00
bel
1364711ee6 Fix unexpected 301 response - revert accidental change 
commit 7e16295a57 (5 Jun 2016)
accidentally reverted commit
commit 455cc3c087
(commit d635fe8e04)
merged with #317
 2016-07-06 22:23:41 +02:00
bel
186d3792b1 Fix possible unaligned access 2016-07-06 22:16:54 +02:00
bel2125
cb084c55f4 Use eventfd instead of pthread_cond - set master back to default 2016-07-05 22:19:47 +02:00
bel2125
45873a5590 Use eventfd instead of pthread_cond (step 8/?) - test ALTERNATIVE_QUEUE in CI tests 2016-07-05 22:10:20 +02:00
bel2125
b4aef8f63c Use eventfd instead of pthread_cond (step 8/?) - test ALTERNATIVE_QUEUE in CI tests 2016-07-05 22:00:51 +02:00
bel
c231cc05a5 Use eventfd instead of pthread_cond (step 7/?) 2016-07-04 22:42:53 +02:00
bel
2ce0e7562a Use eventfd instead of pthread_cond (step 6/?) - test ALTERNATIVE_QUEUE in CI tests 2016-07-04 22:07:21 +02:00
bel
8d166b2d4b Use eventfd instead of pthread_cond (step 5/?) 2016-07-04 21:52:11 +02:00
bel
ff52c41922 Use eventfd instead of pthread_cond (step 4/?) 2016-07-04 21:43:23 +02:00
bel
abdacc98e4 Use eventfd instead of pthread_cond (step 3/?) 2016-07-04 21:33:48 +02:00
bel
fd43dd37b1 Use eventfd instead of pthread_cond (step 2/?) 2016-07-03 22:42:26 +02:00
bel
7023ea5643 Use eventfd instead of pthread_cond (step 1/?) 2016-07-03 22:24:51 +02:00
bel
435c800467 Fix signed/unsigned warning for mg_sleep 2016-07-02 01:11:16 +02:00
bel
a0a5cbce8a It is indeed possible for SSL_write to fail with SSL_ERROR_WANT_READ. 
See http://openssl.6102.n7.nabble.com/SSL-write-fails-td17473.html
 2016-07-02 01:07:13 +02:00
bel
4cf7d820fa Reformat some expressions 2016-07-02 00:32:51 +02:00
bel
202341e3cd Improve SSL error handling and reformat code 2016-07-02 00:30:24 +02:00
bel
0c054e78ed Repeat OpenSSL Read/Write calls if the result is error 2/3 (possible fix for #320) 2016-07-01 22:15:15 +02:00
bel
4abc5be165 Use CMAKE_BUILD_TYPE instead of hard coded "Release" 2016-06-25 00:08:58 +02:00
bel
3ca45d6dae Remove windows specific code from example 2016-06-24 22:23:03 +02:00
bel
3e752feea6 try to fix a curious appveyor build failure that suddenly occured without change 2016-06-24 22:18:40 +02:00
bel
6bc601256a Auto format after merge 2016-06-24 00:19:05 +02:00
bel2125
03b3428fb9 Merge pull request #322 from xtne6f/fix-thread 
Fix thread handling
 2016-06-24 00:04:35 +02:00
xtne6f
51c2afe8f1 Fix and improve pthread_cond in Windows 
Fix possible overrun of waitingthreadhdls in case of pthread_cond_timedwait() with
abstime (fortunately abstime is not used).
Related commit is 087cb745e5 .
Related discussion is https://github.com/sunsetbrew/civetweb/pull/30 .
 2016-06-14 22:03:21 +09:00
xtne6f
faf6f1f3de Remove an useless condition 
This must be an old relic to join threads.
 2016-06-14 19:57:11 +09:00
xtne6f
e071d4691e Fix unlocked broadcast 
Since the purpose of this broadcast is to notify stop_flag, locking is
once needed.
 2016-06-14 19:52:49 +09:00
xtne6f
1c0212798c Preallocate memory for pollfds 
Because master_thread has no proper way to exit when allocation failed.
 2016-06-14 19:39:05 +09:00
xtne6f
c2e419910f Remove listening_ports dynamic array 
"listening_sockets" has duplicate information, so we have only to update
this.
Related commit is 2a4245e379 .
 2016-06-14 19:08:52 +09:00
bel
aec25534c2 If NO_SSL is defined, dlopen is not required 2016-06-12 23:24:57 +02:00
bel
16e335a6eb Some fixes for WinCE 2016-06-12 22:51:06 +02:00
bel
8be52568e8 Add new unit test 2016-06-10 21:47:21 +02:00
bel
e783af1fc3 Add #ifdef to remove function not used if NO_CACHING is not set 2016-06-10 19:39:19 +02:00
bel
150ad4a2e7 Add test for multiline header and do not allow spaces (#318) 2016-06-10 19:33:23 +02:00
bel
7accd96c7e RFC 7230 obsoletes multiline header from RFC 2616 
According to
https://tools.ietf.org/html/rfc7230#section-3.2.4
multiline header can return "400 Bad Request".

See #318
 2016-06-09 22:20:16 +02:00
bel
bade30120f Remove excess lines in test (copy paste error), see #318 2016-06-08 22:26:40 +02:00
xtne6f
d98f9a203a Fix response header injection on ssl-redirect using maliciously encoded URIs 2016-06-06 23:51:11 +02:00
xtne6f
81a72a9c59 Fix parsing start-line for avoiding response header injection 2016-06-06 23:50:42 +02:00
xtne6f
f3449a8af5 More compliant 304 response 2016-06-06 23:49:32 +02:00
xtne6f
bf9ceb86f8 Do not send no-cache header for 304 response 2016-06-06 23:48:59 +02:00
xtne6f
e0175ea806 Resolve merge conflict 2016-06-06 23:48:56 +02:00
xtne6f
1afdde421f Fix lsp_url_encode() 
Dst might be 3 times as large as the input
 2016-06-06 23:43:40 +02:00
xtne6f
97368471d0 Extend mg.write() to return a boolean 2016-06-06 23:43:02 +02:00
xtne6f
9f051f3ddc Fix the size restriction of percent-encoding 2016-06-06 23:41:39 +02:00