Commit Graph

5942 Commits

Author SHA1 Message Date
Niels Dossche
75441d71d8 Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Fix GH-10964: Improve `man` page about the built-in server
  Fix GH-11438: mysqlnd fails to authenticate with sha256_password accounts using passwords longer than 19 characters
2023-08-03 20:32:30 +02:00
Alexandre Daubois
997a36750b Fix GH-10964: Improve man page about the built-in server
Closes GH-11857.
2023-08-03 20:29:13 +02:00
Niels Dossche
272dc9a09c Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Fix GH-11716: cli server crashes on SIGINT when compiled with ZEND_RC_DEBUG=1
2023-07-21 15:31:20 +02:00
Niels Dossche
af77d3b8da Fix GH-11716: cli server crashes on SIGINT when compiled with ZEND_RC_DEBUG=1
Closes GH-11757.
2023-07-21 15:23:56 +02:00
Niels Dossche
4a28cc725a Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Fix GH-9669: phpdbg -h options doesn't list the -z option
2023-07-07 19:17:35 +02:00
Adam Saponara
119b062169 Fix GH-9669: phpdbg -h options doesn't list the -z option
Adds `-z` flag in phpdbg output.

Closes GH-9713.
2023-07-07 19:16:32 +02:00
nielsdos
b9bf9ee2cb Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Fix interrupted CLI output causing the process to exit
2023-06-23 17:42:45 +02:00
nielsdos
1111a9517b Fix interrupted CLI output causing the process to exit
When writing the output in the CLI is interrupted by a signal, the
writing will fail in sapi_cli_single_write(), causing an exit later in
sapi_cli_ub_write(). This was the other part of the issue in GH-11498.
The solution is to restart the write if an EINTR has been observed.

Closes GH-11510.
2023-06-23 17:39:04 +02:00
Jakub Zelenka
6877b4f110
Merge branch 'PHP-8.1' into PHP-8.2 2023-06-15 17:37:09 +01:00
Jakub Zelenka
9b18466396
FPM: Add "pcntl" when running another test depending on pcntl 2023-06-15 17:36:00 +01:00
Mikhail Galanin
7ade242e28
sapi/fpm: add "pcntl" when running test depending pcntl_sigprocmask()
If "pcntl" is built as a shared module, the extension will not
load automatically when we spawn the FPM
2023-06-15 17:32:02 +01:00
Ilija Tovilo
0bf8ecf9b3
Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Fix bug GH-11246 cli/get_set_process_title
2023-05-31 17:10:17 +02:00
James Lucas
c6ae7a55b7
Fix bug GH-11246 cli/get_set_process_title
Fail to clobber_error only when the argv is a non-contiguous area
Don't increment the end_of_error if a non-contiguous area is encountered in environ

Closes GH-11247
2023-05-31 17:08:37 +02:00
Jakub Zelenka
3769505c87
Merge branch 'PHP-8.1' into PHP-8.2 2023-05-19 16:30:36 +01:00
Jakub Zelenka
aa061cd40b
Fix FPM status json encoded value test
Closes GH-11276
2023-05-19 16:20:21 +01:00
Ilija Tovilo
69d41cc7fc
Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  [skip ci] Mark frequently failing fpm test as XFAIL
2023-05-17 17:42:39 +02:00
Ilija Tovilo
e31fe111a5
[skip ci] Mark frequently failing fpm test as XFAIL
Reported here: https://github.com/php/php-src/pull/11050#issuecomment-1546990346
2023-05-17 17:42:05 +02:00
Jakub Zelenka
1129046482
Merge branch 'PHP-8.1' into PHP-8.2 2023-05-14 12:28:52 +01:00
Jakub Zelenka
4294e8d448
FPM: Fix memory leak for invalid primary script file handle
Closes GH-11088
2023-05-14 12:27:38 +01:00
Jakub Zelenka
90553af15c
Merge branch 'PHP-8.1' into PHP-8.2 2023-05-13 18:53:35 +01:00
Jakub Zelenka
5e64ead64a
Fix bug #64539: FPM status - query_string not properly JSON encoded
Closes GH-11050
2023-05-13 18:43:30 +01:00
Jakub Zelenka
bda28eb649
Merge branch 'PHP-8.1' into PHP-8.2 2023-05-13 14:14:23 +01:00
Jakub Zelenka
102953735c
Fix GH-10461: Postpone FPM child freeing in event loop
This is to prevent after free accessing of the child event that might
happen when child is killed and the message is delivered at that same
time.

Also fixes GH-10889 and properly fixes GH-8517 that was not previously
fixed correctly.
2023-05-13 14:11:42 +01:00
Jakub Zelenka
0b4f2fc9e8
Merge branch 'PHP-8.1' into PHP-8.2 2023-04-07 12:29:23 +01:00
Jakub Zelenka
ebb3213f79
Add FPM FCGI env Apache handler UDS test 2023-04-07 12:28:42 +01:00
Jakub Zelenka
1ca1858f31
Merge branch 'PHP-8.1' into PHP-8.2 2023-04-06 16:00:48 +01:00
Jakub Zelenka
15802dfc62 Add FPM FastCGI env var test for Apache without path info fix 2023-04-06 15:59:34 +01:00
Ilija Tovilo
9b15537e9a
Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Unparallelize IO heavy tests
2023-03-27 16:34:11 +02:00
Ilija Tovilo
e1ec67acd6
Unparallelize IO heavy tests
Alternative to GH-10892. This is somewhat unfortunate since these are also the
slow tests. I'm also not sure if this actually helps.

Closes GH-10953
2023-03-27 16:33:36 +02:00
Jakub Zelenka
2b0d29bf17
Merge branch 'PHP-8.1' into PHP-8.2 2023-03-19 13:17:09 +00:00
Jakub Zelenka
0d4d471847
Fix bug #74129: Incorrect SCRIPT_NAME with apache ProxyPassMatch
This happens when there are spaces are in the path info. The reason is
that Apache decodes the path info part in the SCRIPT_NAME as per CGI
RFC. FPM tries to strip path info from the SCRIPT_NAME but the
comparison is done against SCRIPT_FILENAME which is not decoded. For
that to work we have to decode it before comparison if there is any
encoded character.

Closes GH-10869
2023-03-19 13:11:14 +00:00
Niels Dossche
e44b1a916d Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Destroy file_handle in fpm_main
2023-03-18 21:00:21 +01:00
Niels Dossche
f30e71cbde Destroy file_handle in fpm_main
If it's not in the CG(open_files) list, we need to destroy the file
handle ourselves.

Co-authored-by: Jakub Zelenka <bukka@php.net>

Closes GH-10707.
2023-03-18 20:54:46 +01:00
Jakub Zelenka
4ac12006ae
Merge branch 'PHP-8.1' into PHP-8.2 2023-03-17 14:56:45 +00:00
Jakub Zelenka
7d987ebbbf Fix FPM tester $scriptName logic 2023-03-17 14:55:51 +00:00
Jakub Zelenka
e9bbddcdaf
Merge branch 'PHP-8.1' into PHP-8.2 2023-03-17 10:45:14 +00:00
Jakub Zelenka
38d2e7ea9a
Test FPM FCGI envs with path info fix for Apache proxy pass 2023-03-17 10:42:16 +00:00
Jakub Zelenka
8cf621e0e4
Test FPM FCGI envs with path info fix for Apache proxy handler 2023-03-17 10:42:16 +00:00
Jakub Zelenka
b53b0ac2ea
Test FPM FCGI envs with path info fix for Apache proxy balancer 2023-03-17 10:42:15 +00:00
Jakub Zelenka
92d2cd5cb8
Test FPM FCGI envs without path info fix for custom source 2023-03-17 10:42:15 +00:00
Jakub Zelenka
3125155b5d
Add extra option to FPM tester for handling script filename 2023-03-17 10:42:14 +00:00
Ilija Tovilo
70c29f84b1
Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  [skip ci] Skip upload_2G.phpt on Cirrus
2023-03-15 02:11:49 +01:00
Ilija Tovilo
84e7d4a3c8
[skip ci] Skip upload_2G.phpt on Cirrus 2023-03-15 02:11:11 +01:00
Niels Dossche
6a44705f4e Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Fix GH-10611: fpm_env_init_main leaks environ
2023-03-11 16:35:11 +01:00
Niels Dossche
71c63723c8 Fix GH-10611: fpm_env_init_main leaks environ
Closes GH-10618.
2023-03-11 16:34:41 +01:00
Ilija Tovilo
f415ba60ab
Disable asan instrumentation for phpdbg_watchpoint_userfaultfd_thread
On gcc. It reports a false positive stack-overflow.

Closes GH-10818
2023-03-10 00:13:27 +01:00
Niels Dossche
7682868dd1 Merge branch 'PHP-8.1' into PHP-8.2
* PHP-8.1:
  Fix GH-8065: opcache.consistency_checks > 0 causes segfaults in PHP >= 8.1.5 in fpm context
  Fix GH-8646: Memory leak PHP FPM 8.1
2023-03-07 20:26:57 +01:00
Niels Dossche
ff62d117a3 Fix GH-8646: Memory leak PHP FPM 8.1
Fixes GH-8646
See https://github.com/php/php-src/issues/8646 for thorough discussion.

Interned strings that hold class entries can get a corresponding slot in map_ptr for the CE cache.
map_ptr works like a bump allocator: there is a counter which increases to allocate the next slot in the map.

For class name strings in non-opcache we have:
  - on startup: permanent + interned
  - on request: interned
For class name strings in opcache we have:
  - on startup: permanent + interned
  - on request: either not interned at all, which we can ignore because they won't get a CE cache entry
                or they were already permanent + interned
                or we get a new permanent + interned string in the opcache persistence code

Notice that the map_ptr layout always has the permanent strings first, and the request strings after.
In non-opcache, a request string may get a slot in map_ptr, and that interned request string
gets destroyed at the end of the request. The corresponding map_ptr slot can thereafter never be used again.
This causes map_ptr to keep reallocating to larger and larger sizes.

We solve it as follows:
We can check whether we had any interned request strings, which only happens in non-opcache.
If we have any, we reset map_ptr to the last permanent string.
We can't lose any permanent strings because of map_ptr's layout.

Closes GH-10783.
2023-03-07 20:16:17 +01:00
David Carlier
4c2db35b05 Merge branch 'PHP-8.1' into PHP-8.2 2023-02-27 19:48:13 +00:00
Niels Dossche
0f21cbc57c Fix GH-10715: phpdbg heap buffer overflow -- by misuse of the option "--run"
Fixes GH-10715

When a string starting with a NUL character is passed to
phpdbg_vprint(), the vasprintf() will return that 0 characters have been
printed. This causes msglen == 0. When phpdbg_process_print() is called
with a message of length 0, the -1 to check for '\n' will perform an out
of bounds read. Since nothing is printed anyway for msglen == 0, it
seems best to just skip the printing routine for this case.

Closes GH-10720.
2023-02-27 19:47:55 +00:00