Access denied | www.archivum.info used Cloudflare to restrict access
Please enable cookies.
What happened?
The owner of this website (www.archivum.info) has banned your access based on your browser's signature (9883-ua98).31257 & java.endorsed.dirs is not used when JSP compilation is forked
ASF Bugzilla & Bug&31257
- java.endorsed.dirs is not used when JSP compilation is forked
java.endorsed.dirs is not used when JSP compilation is forked
Unclassified
Windows XP
Tomcat Developers Mailing List
05:10 UTC by James DeFelice
Attachments
This is ASF Bugzilla: the Apache Software Foundation bug system. In case
of problems with the functioning of ASF Bugzilla, please contact
Please Note: this e-mail address is only for reporting problems
with ASF Bugzilla. Mail about any other subject will be silently硬币餐叉minute-to-win-it-get-forked_腾讯视频
三倍流畅播放
1080P蓝光画质
新剧提前看
1080P蓝光画质
纯净式无框播放器
三倍流畅播放
扫一扫 手机继续看
下载需先安装客户端
{clientText}
客户端特权:
3倍流畅播放
当前播放至 {time}
扫一扫 手机继续看
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要
副标题要不要Keyboard Shortcuts?
Next menu item
Previous menu item
Previous man page
Next man page
Scroll to bottom
Scroll to top
Goto homepage
Goto search(current page)
Focus search box
Change language:
Brazilian Portuguese
Chinese (Simplified)
sem_get & Get a semaphore id
resource sem_get
( int $key
[, int $max_acquire = 1
[, int $perm = 0666
[, int $auto_release = 1
A second call to sem_get() for the same key
will return a different semaphore identifier, but both
identifiers access the same underlying semaphore.
max_acquire
The number of processes that can acquire the semaphore simultaneously
is set to max_acquire.
The semaphore permissions. Actually this value is
set only if the process finds it is the only process currently
attached to the semaphore.
auto_release
Specifies if the semaphore should be automatically released on request
Returns a positive semaphore identifier on success, or FALSE on
When using sem_get() to access a semaphore created
outside PHP, note that the semaphore must have been created as a set of 3
semaphores (for example, by specifying 3 as the nsems
parameter when calling the C semget() function),
otherwise PHP will be unable to access the semaphore.
- Acquire a semaphore
- Release a semaphore
- Convert a pathname and a project identifier to a System V IPC key
Actually it looks like the semaphore is automatically released not on request shutdown but when the variable you store it's resource ID is freed. That is a very big difference.
For those that encounter strange behavior in using sem_acquire() on resources generated by sem_get(). Have a look at& sem_get()'s 4th parameter auto_release. It allows multiple acquisitions through reassignments to resource variables../multi.acquire.php&?phpclass Sem {& private $key = null;& private $res = null;& public function __construct() {& & $this-&key = ftok(".",".");& & $this-&set_res();& & $this-&acquire();& }& public function set_res() {& & $this-&res = sem_get($this-&key, 1, 0600, 1);& }& public function acquire() {& & echo "acquired='".sem_acquire($this-&res,true)."'\n";& }}$s = new Sem();$s-&set_res();$s-&acquire();?&$ php multi.acquire.phpacquired='1'acquired='1'To avoid reacquiring by default set sem_get()'s parameter auto_release to 0 or check if your resource variable is already set, e.g. by using is_null().
It is possible to create an "infinite" amount of semaphores when setting $key = 0.Run sem_get multiple timesphp & sem_get(0,0);and check the output of$ ipcs -s------ Semaphore Arrays --------key& & & & semid& & & owner& & & perms& & & nsems& && 0x7952& & user& & && 666& & & & 3& & & && 0x0721& & user& & && 666& & & & 3As you can see there were multiple semaphores set up with key 0.For any other integer sem_get works as expected. It returns another resource id pointing to the semaphore previously created and does not create another semaphore.
Watch out when you use fileinode() to get a unique semaphore key (as suggested in some comment on this or a related function) in conjunction with version control software: It seems, for example, SVN will change the inode. Using such a file will leave you with your mutex not working reliably and your system's semaphore pool being filled until further attempts to get a semaphore will fail. Use ipcs and ipcrm commands from linux-util-ng (on most distros probably) to examine/fix related problems.
with gentoo php5 you will need to add the USE flag :sysvipcsee : and also :
Heh, actually the above comment I added is not technically correct, it was more of an idea to display the function.$SHM_KEY = ftok("/home/joeldg/homeymail/shmtest.php", 'R');$shmid = sem_get($SHM_KEY,
| IPC_CREAT);$data = shm_attach($shmid, 1024);// we now have our shm segment// lets place a variable in thereshm_put_var ($data, $inmem, "test");// now lets get it back. we could be in a forked process and still have// access to this variable.printf("shared contents: %s\n", shm_get_var($data, $inmem));shm_detach($data);
&?// thanks to// $SHM_KEY = ftok("/home/joeldg/homeymail/shmtest.php", 'R');$shmid = sem_get($SHM_KEY,
| IPC_CREAT);$data = shm_attach($shmid, 1024);$data = "test";printf("shared contents: %s\n", $data);shm_detach($data);?&
Implementation of a read-write semaphore in PHP:&?phpclass rw_semaphore {& & & & & & const READ_ACCESS = 0;& & const WRITE_ACCESS = 1;& & & & & & private $mutex;& & & & private $resource;& & & & private $writers = 0;& & & & private $readers = 0;& & public function __construct() {& & & & $mutex_key = ftok('/home/cyrus/development/php/sysvipc/rw_semaphore.php', 'm');& & & & $resource_key = ftok('/home/cyrus/development/php/sysvipc/rw_semaphore.php', 'r');& & & & & & & & $this-&mutex = sem_get($mutex_key, 1);& & & & $this-&resource = sem_get($resource_key, 1);& & & & & & }& & & & public function __destruct() {& & & & sem_remove($this-&mutex);& & & & sem_remove($this-&resource);& & }& & & & private function request_access($access_type = self::READ_ACCESS) {& & & & & & if ($access_type == self::WRITE_ACCESS) {& & & & & & sem_acquire($this-&mutex);& & & & & & & & & & & & $this-&writers++;& & & & & & & & & & & & sem_release($this-&mutex);& & & & & & & & & & & & sem_acquire($this-&resource);& & & & } else {& & & & & & & & & & & & sem_acquire($this-&mutex);& & & & & & & & & & & & if ($this-&writers & 0 || $this-&readers == 0) {& & & & & & & & & & & & & & & & sem_release($this-&mutex);& & & & & & & & & & & & & & & & sem_acquire($this-&resource);& & & & & & & & & & & & & & & & sem_acquire($this-&mutex);& & & & & & & & & & & & & & }& & & & & & $this-&readers++;& & & & & & & & & & & & sem_release($this-&mutex);& & & & }& & }& & & & private function request_release($access_type = self::READ_ACCESS) {& & & & if ($access_type == self::WRITE_ACCESS) {& & & & & & sem_acquire($this-&mutex);& & & & & & & & & & & & $this-&writers--;& & & & & & & & & & & & sem_release($this-&mutex);& & & & & & sem_release($this-&resource);& & & & } else {& & & & & & sem_acquire($this-&mutex);& & & & & & & & & & & & $this-&readers--;& & & & & & & & & & & & if ($this-&readers == 0)& & & & & & & & sem_release($this-&resource);& & & & & & & & & & & & sem_release($this-&mutex);& & & & }& & }& & & & public function read_access() { $this-&request_access(self::READ_ACCESS); }& & & & public function read_release() { $this-&request_release(self::READ_ACCESS); }& & & & public function write_access() { $this-&request_access(self::WRITE_ACCESS); }& & & & public function write_release() { $this-&request_release(self::WRITE_ACCESS); }& & }?&
Be aware that there is no way to ensure that you have exclusive access to a lock, despite setting max_acquire=1.In example,&?$fp = sem_get(fileinode('lock_file', 100);sem_acquire($fp);$fp2 = sem_get(fileinode('lock_file', 1);sem_acquire($fp2);?&This will not block on the second sem_aquire.& Therefore, if you have functions or processes that utilize shared locks (&1 max_acquire) you will still need to provide a seperate lock mechanism (ie flock) for write access, making the sem_ functions useless.Some more info, in flock, each reference to the lock file has it's own options (can be shared exclusive blocking non blocking etc), but apparently php's sem functions only support these options per semaphore, not per semaphore-reference.
A very simple to introduce semaphore...
&?php
$SEMKey = "123456" ;
$seg = sem_get( $SEMKey, 2, 0666, -1) ;
if ( $argv[1]=="remove" ) {
& & sem_remove($seg);
}
echo "Try to acquire ..."
sem_acquire($seg);
echo "Acquired...\n" ;
echo "Press Any Key to continue...\n";
$fh = fopen("php://stdin", "r");
$a = fgets( $fh);
fclose($fh);
sem_release($seg);
?&Re: [DISCUSS] Get off Calcite Forked Version
drill-dev mailing list archives
Message view
Julian Hyde &jh...@apache.org&
Re: [DISCUSS] Get off Calcite Forked Version
Going forward, I think you should start a "calcite first" policy --
get changes accepted into Calcite, with tests of course, then
cherry-pick into Drill's branch. Recent changes like "Allow a MAP
literal type." should have done that. And especially changes where you
could just parameterize planner rule. In other words, stop digging the
hole deeper.
I think you should identify and do the easy ones ASAP.
Rather than creating a branch for each of these contributions, if you
prefer you could create a pull request of the whole branch I think
that individual JIRAs can just reference a commit id that a Calcite
committer can cherry-pick.
Now some remarks on specific changes.
Why do you need "Add new method to ViewExpander interface to allow
passing SchemaRoot."? Most ViewExpander implementations have a root
schema. Note that the return type of that method has changed from
RelNode to RelRoot.
"Match Logical Rel in ReduceExpressionRule." should be easy now we've
parameterized the rules, and maybe you can get rid of "Enable
inheriting from previously singleton filter and calc constant
reduction rules.".
"Disable the nullability cast checking in Filter." is worth discussing
on the Calcite list.
In "Return validated row type and validated SqlNode when call
Planner.validate()." you should return a Pair&SqlNode, RelDataType&.
To do "Use case-insensitive comparison when creating output row type
of a JoinRel." properly we require a new config parameter saying
whether internal field names are case-sensitive.
Since "Don't use 'SumEmptyIsZero' (SUM0) window aggregate until
CALCITE-777 is fixed." requires a fix to CALCITE-777 please contribute
a fix for CALCITE-777! :)
Why do you need "Make public the constructor of SqlSumEmptyIsZeroAggFunction,"?
You don't need "Make certain push project constructors protected such
that derived classes can access them." now we've parameterized
ProjectFilterTransposeRule.
"DRILL-1455: Add return type-inference strategy for arithmetic
operators when one of the arguments is ANY type." needs some
discussion. What if both are ANY? How are other implementations of the
operators (e.g. Calcite, Phoenix) going to implement the operator now
that Drill's lax validation policy has allowed it in?
I can't see yet how the [StarColumn] changes could make it in. Will
require some discussion. Note that CALCITE-546 has broken some of your
assumptions.
On Tue, Nov 10, 2015 at 12:27 PM, Jinfeng Ni &jni@apache.org& wrote:
& Currently, Drill maintains a forked version of Calcite. Because of
& this, it has been a big headache to rebase onto the latest Calcite
& master branch every time when Calcite makes a new release.
& During today's Drill hangout, we discuss ideas around how to get Drill
& off the forked Calcite, by either pushing back patches in the forked
& version to Calcite master branch, or trying to see if those patches
& are still required for Drill to work properly.
& There are 49 commits in forked Calcite whose base is Calcite 1.4.0
& release [1]. I put a spreadsheet for the current status of those 49
& commits, as far as I know [2]. Half of them are the patches that were
& cherry-picked from Calcite master branch,
9 are just for version
& change, and the rest requires either push back or re-work in Drill
& I think we probably have to spend considerable effort to get this done
& ASAP. Otherwise, it will continue to be an issue for Drill.
& There are two separate work to be done:
& 1. Continue to rebase forked version onto latest Calcite master
& (Currently it's 1.5.0)
& 2. Push the patches in fork to Calcite master. Once all patches are
& in, we could finally get rid of fork and rebasing efforts.
& The spreadsheet has a column for sign-up.
& Thoughts?
& (let me know if you have problem accessing the spreadsheet).
& [1] https://github.com/dremio/calcite/tree/DrillCalcite1.4.0
& [2] https://docs.google.com/spreadsheets/d/1Z0J5aMCBfqq_9SEl-LXsi_B8Ahaosygqke4cH6pPwmo/edit#gid=0
(inline, None, 4001 bytes)}