Variants included 'in <version>', 'as of <version>' and just the
version number.
Some @deprecated annotations do not have the version number at all,
I want to hunt them down separately.
Change-Id: I8208c6097098f4735d4f51bc42254675f1f27f6d
This patch replaces:
<!-- Served by mw1069 in 0.976 secs. -->
With:
<script>mw.config.set({"wgBackendResponseTime":976,"wgHostname":"mw1069"});</script>
In the default HTML output of MediaWiki.
While the latter is a nearly twice as long, it is almost as readable for human
beings, while being substantially easy to get via JavaScript.
To get the values from the comment, you have to do something like:
var comments, comment, hostname, duration;
comments = $.grep( document.body.childNodes, function ( el ) {
return el.nodeType === 8
} );
comment = comments.length
? comments.pop().nodeValue.match( /(\S+) in ([\d.]+)/ ).slice( 1 )
: [ null, null ];
hostname = comment[0];
respTime = parseFloat( comment[1] );
On the other hand, to get the values from the JavaScript code, you can simply:
var hostname = mw.config.get( 'wgHostname' );
var respTime = mw.config.get( 'wgBackendResponseTime' );
I believe that the ability to parse the number easily via JavaScript will make
it easier to include with other client-side measurements as part of reports on
site performance as experienced by users.
Change-Id: I895cd03f0968815484ff8cda4b23cc602ac555f0
And deprecate $wgProfileOnly in the same time.
This has the advantage of allowing profiler output to be separated
from the main debug log file; or even be completely disabled while
keeping the other debugging messages.
Also updated the checks in wfLogProfilingData() to detect the cases
where the output would not be sent anywhere to not execute the
last part of the method which would be useless otherwise.
Backward compatibility with installations having $wgProfileOnly
set to true is kept by moving the log file from $wgDebugLogFile
to $wgDebugLogGroups['profileoutput'] in Setup.php in that case.
Change-Id: I7b35195e527dfa7978b710126ed4599e75dab46b
I mainly fixed PHPDoc comments that are highlighted by PHPStorm
as being wrong or incomplete.
* Fixed some wrong inline documentations, e.g. wrong @return types.
* Used the style suggested by the Doxygen and phpDocumentor
documentations: @param type [$optional,...]
* Doxygen and phpDocumentor suggest the order: @param type $var
* Avoid unclear "varargs" and "mixed" if possible.
* Lowercased "array" and such, they are no classes.
Change-Id: Iff16e05fed6652cefec29e3d254807f52d84c9ab
Since Iffba121a99 (00b7f76) with the removal of wfHttpOnlySafe(),
session cookie's parameters are based only on configuration
settings, so there is no point to spam the "cookie" log group
with predicitible values.
Change-Id: I8b1cdea929cefc32dd8b01c2ecbf2d76bb64189f
I found two calls to wfLogDBError() that do not add a new line
at the end of the message. So instead of adding them to that
entries, I changed wfLogDBError() to automatically put it on
icoming messages; as for wfDebugLog().
Change-Id: Id014b5827a0aeef6873ebf08d78f0a3d7581d63b
This hack was added in r34083 / 6b16f44108 to support IE for Mac.
That browser is no longer supported, and no additional user-agent
strings have been added in WMF configuration.
Change-Id: Iffba121a9964e2ad387fad8827ddfd8dabcbd12e
- The parameter is now a string, making is more understandable than
boolean values
- It takes the same values in both wfDebug() and wfDebugLog() (except
for 'private' which is only used in the latter)
- This adds a new possibility to wfDebugLog() to log the message either
on the specific log or the general one, but not to the debug toolbar
- Old boolean values are still recognised for backward compatibility
- Also send the messages passed to wfDebugLog() to the debug toolbar
when they are written to a specific log and not restricted to logs
- Updated the calls of and wfDebug() and wfDebugLog() with the last
parameter to change it into a string
- Renamed MWDebug::sendWarning() to MWDebug::sendMessage() and added
$group parameter to it; will not break anything since that method
is marked as private
- Changed the call to wfDebug() from MWDebug::sendMessage() to use
wfDebugLog() with 'log' as thrid parameter, so that those messages
can be logged separately from the main log and they don't show up
a second time on the "debug log" tab of the debug toolbar
Change-Id: I1be09d4c1d3408ed5b26a5db02691c17c0ec0926
Some uses remained in Gerrit extensions, and patches have been submitted
to replace the usages. See below for details.
* in_string (deprecated in 1.21)
* wfGenerateToken (deprecated in 1.20) (had 2 uses)
* wfStreamFile (deprecated in 1.19) (had 1 use)
* wfUILang (deprecated in 1.18)
Change-Id: Ic67a968da8f37daf44503b98d00710eb99180805
Kept wfGetIP() for now because it is still in use in the SecurePoll
extension and in WMF configuration.
Change-Id: Id412e37743f89bb67eb329a2f475f8496cf7b006
The Line continuation Coding conventions prefers the closing parenthesis
on the same line than the beginning curly braces. This is done for ifs
and functions.
Also move some boolean operator from the end of a line to the beginning
and changed some indentation to make the condition hopefully better
readable.
Change-Id: Id0437b06bde86eb5a75bc59eefa19e7edb624426
- Removed double spaces
- Added space after if/switch/foreach
- Removed space on elseif
- Added space around parentheses
- Added newline at end of file
- Removed space before semicolon at end of line
Change-Id: Id40b87e04786c6111e6686d7f7eea1e588bdf37d
The use of static server detection outside of its intended use case
(i.e. at the start of DefaultSettings.php), for example in r93258, was
an architectural error. Every other bit of information about the web
request in non-setup code comes from non-static methods of WebRequest,
which allows the request object to be meaningfully replaced or
subclassed. The situation became increasingly ridiculous as more
callers of WebRequest::detectProtocol() were introduced. Two of the
callers were calling it non-statically! I suppose they had the right
idea, in a way.
Using a non-static call allows caching, which is a nice additional
benefit.
WebRequest::detectProtocolAndStdPort() was introduced in r93258 as part of
the introduction of WebRequest::detectProtocol(). It was basically
useless. Grep indicates there was only one caller in core and WMF
deployed extensions, and it is patched here.
Change-Id: Ia0a61e98fbff7a46ceaeebcb02236e5eac3df0e1
Adds release notes and fixes two inaccurate comments added in Ib0ece5d5d
("Configurable log sampling via $wgDebugLogGroups").
Change-Id: I224a0319b7432300d3cb6f4d14e61ea9d36ee228
Make it possible to configure a log-group-specific sampling factor by allowing
$wgDebugLogGroups values to be associative arrays with 'destination' & 'sample'
keys.
The immediate use-case I have in mind is the memcached log, which is too
verbose to enable in production. We could sample on the receiving end, but
doing so would not help MediaWiki, which would still be on the hook for
processing and dispatching every single record.
Idea was vetted by Dr. Manhattan:
<TimStarling> I think you could have each element in $wgDebugLogGroups be an
associative array
<TimStarling> e.g. array( 'sample' => 1000, 'target' => "$host:$port")
<TimStarling> that would allow for more features to be added to it in
future
Change-Id: Ib0ece5d5d9c2aadef13b77ac38995b0e47ac086f
The former is only really used to refer to the older hphp compiler.
Now all docs pretty much exclusively are referring to HHVM and
that's the project's name going forward.
While we're here, swap HPHP_VERSION for HHVM_VERSION. Either work,
but the latter is more correct.
Only affected extension (in Git or SVN) is GWToolset, fixing in
I3a490ef.
Change-Id: I81e4ce5d14e344e67045050261f91b3c0159f222
Allow limit.sh to log its errors to an MW debug log channel, by opening
a separate FD for private communication.
Ensure that the log FD is always closed (3>&-) before executing a
subprocess, so that MW will not hang waiting for background processes
to close the log FD. This means using a fixed FD number, since the bash
syntax for closing a file requires a literal FD number.
The "exec" debug channel is now intended for production monitoring. In
addition to errors from limit.sh, it also records when a subprocess is
terminated by a signal.
The case where stream_select() returns false was tested by patching PHP
to inject EINTR or EBADF into errno.
When wfShellExec() is used with a memory cgroup, and the cgroup is missing, log and continue with no cgroup instead of immediately exiting.
Bug: 55709
Change-Id: Ie40befe9c0d00c9a0ddb01077df4afb774d17e15
While reviewing Ia5a7661c I happened to stumble in to the dark labyrinth
of PHP's version_compare() function. This comment attempts to leave
bread crumbs behind for the next unwary adventurer who stumbles into the
same damp cave.
The tricky bit is that when '1.22wmf22' is compared with '1.22'
internally it is as though you are comparing '1.22.wmf.22' with
'1.22.0.0'. By the component-wise comparison then performed 'wmf' is
considered the lowest priority value that could be in the third
position. When using '1.22c' for the required version the 'c' is given
the same lowest priority as 'wmf'.
Change-Id: I58aeb8dc6b1ff31897d72eb44904d96487ad3653
* Separate code from data. $2 in limit.sh is an option list.
* Don't duplicate stderr on linux when limit.sh is not used, unless
duplicateStderr is specified
Change-Id: I8e6e74810864830c12af627bfbe75e74ba34e3e3
Per the bug report, the 'Debug' hook was triggering an infinite loop when
wgDebugFunctionEntry is enabled. The Debug hook is used if an extension wants
to stop a debug message from being sent out. Ideally the wfDebug and related
functions should be as low-level and avoid calling other code as much as
possible to avoid situations like this.
Bug: 55818
Change-Id: I679782489b683503fc624cfea3c7ad72a989b005
wfDl() is a wrapper around dl(), which is an evil function and
basically only works from the command line of Zend. Luckily
no extension has ever used this thing, so let's just remove it
outright.
For comparison, here's a list of places it does not work:
- hhvm
- php as apache module
- php compiled with zts support
- safe_mode
- Basically any shared host that cares about security
Most callers are using it to check for extension support and are
actually failing gracefully when wfDl() returns false. In these
places we're just going to use extension_loaded().
While we're at it, clean up some of the test skip logic in the
media tests so we can bail as early as possible if we know we
can't complete the test.
This also immediately removes $wgLoadFileinfoExtension. It's been
enabled by default since 5.3 and falls back gracefully when the
support isn't available.
Change-Id: Ieb430dfc74483731dde51d6e20fa700d641ba1f4
This adds an option to wfShellExec (and convenience function
wfShellExecWithStderr), to make sure all stderr is duplicated
to stdout. The previous method of doing this was to include
2>&1 on the command line. However this did not redirect errors
from limit.sh (For example cgroups not set up, or if a command
reached the file size limit set by ulimit).
Not sure if this is the best approach, but it seems to work well,
and compared to most other approaches I considered, actually gets
the ulimit errors redirected too.
Currently some files fail to render with no error whatsoever,
hopefully this patch will make what went wrong more obvious.
Also fix a comment in wfShellExec that was incorrect (trailing \n),
and make the initial value of the return value variable be 200, so
if there's ever a bug in php where its not being set properly, it
would be immediately obvious what is happening.
Bug: 53824
Change-Id: I833aeb3ab9da726ecb97331369ea187daad7e795
