Bug report for Taglibs [2013/09/22]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=Critical REG=Regression MAJ=Major | | | | MIN=Minor NOR=NormalENH=Enhancement TRV=Trivial | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |38193|Ass|Enh|2006-01-09|[RDC] BuiltIn Grammar support for Field | |38600|Ass|Enh|2006-02-10|[RDC] Enable RDCs to be used in X+V markup (X+RDC)| |42413|New|Enh|2007-05-14|[PATCH] Log Taglib enhancements | |46052|New|Nor|2008-10-21|SetLocaleSupport is slow to initialize when many l| |48333|New|Enh|2009-12-02|TLD generator | +-+---+---+--+--+ | Total5 bugs | +---+ - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Bug report for Tomcat 8 [2013/09/22]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=Critical REG=Regression MAJ=Major | | | | MIN=Minor NOR=NormalENH=Enhancement TRV=Trivial | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |51497|New|Enh|2011-07-11|Use canonical IPv6 text representation in logs| |53737|Opn|Enh|2012-08-18|Use ServletContext.getJspConfigDescriptor() in Jas| |53930|New|Enh|2012-09-24|allow capture of catalina stdout/stderr to a comma| |54095|New|Enh|2012-11-03|[patch] support gzipped versions of static resourc| |54503|New|Enh|2013-01-29|SAML2 based single sign on| |54700|New|Enh|2013-03-15|Improvement: Add support for system property to sp| |54741|New|Enh|2013-03-22|Add org.apache.catalina.startup.Tomcat#addWebapp(S| |55006|New|Enh|2013-05-22|Add http proxy support for ClientEndpoint using sy| |55212|New|Enh|2013-07-08|Jasper does not validate XSD-based configuration f| |55243|New|Enh|2013-07-11|Add special search string for nested roles| |55249|New|Nor|2013-07-12|JspC compiles tag files even if compile options is| |55251|New|Nor|2013-07-12|JspC fails silently if directory for generated web| |55252|New|Enh|2013-07-12|Separate Ant and command-line wrappers for JspC | |55383|New|Enh|2013-08-07|Improve markup and design of Tomcat's HTML pages | |55479|New|Enh|2013-08-24|JSR 196 (JASPIC) support in Tomcat| |55508|Opn|Nor|2013-08-30|TldLocationsCache not recognising custom TLDs in W| |9|New|Enh|2013-09-14|UserDatabaseRealm enhacement: may use local JNDI | +-+---+---+--+--+ | Total 17 bugs | +---+ - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Bug report for Tomcat Connectors [2013/09/22]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=Critical REG=Regression MAJ=Major | | | | MIN=Minor NOR=NormalENH=Enhancement TRV=Trivial | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |34526|Opn|Nor|2005-04-19|Truncated content in decompressed requests from mo| |35959|Opn|Enh|2005-08-01|mod_jk not independant of UseCanonicalName| |43303|New|Enh|2007-09-04|Versioning under Windows not reported by many conn| |43968|Inf|Enh|2007-11-26|[patch] support ipv6 with mod_jk | |44290|Inf|Nor|2008-01-24|mod_jk/1.2.26: retry is not useful for an importan| |44349|Inf|Maj|2008-02-04|mod_jk/1.2.26 module does not read worker.status.s| |44379|New|Enh|2008-02-07|convert the output of strftime into UTF-8 | |44454|New|Nor|2008-02-19|busy count reported in mod_jk inflated, causes inc| |44571|New|Enh|2008-03-10|Limits busy per worker to a threshold | |45063|New|Nor|2008-05-22|JK-1.2.26 IIS ISAPI filter issue when running diff| |45313|New|Nor|2008-06-30|mod_jk 1.2.26 apache 2.2.9 static compiled on so| |46337|New|Nor|2008-12-04|real worker name is wrong | |46767|New|Enh|2009-02-25|mod_jk to send DECLINED in case no fail-over tomca| |47327|New|Enh|2009-06-07|remote_user not logged in apache logfile | |47617|Inf|Enh|2009-07-31|include time spent doing ajp_get_endpoint() in err| |47678|New|Cri|2009-08-11|Unable to allocate shared memory when using isapi_| |47714|New|Cri|2009-08-20|Reponse mixed between users | |47750|New|Maj|2009-08-27|Loss of worker settings when changing via jkstatus| |47795|New|Maj|2009-09-07|service sticky_session not being set correctly wit| |47840|Inf|Min|2009-09-14|A broken worker name is written in the log file. | |48191|New|Maj|2009-11-13|Problem with mod_jk 1.2.28 - Can not render up the| |48460|New|Nor|2009-12-30|mod_proxy_ajp document has three misleading portio| |48490|New|Nor|2010-01-05|Changing a node to stopped in uriworkermap.propert| |48513|New|Enh|2010-01-09|IIS Quick setup instructions | |48564|New|Nor|2010-01-18|Unable to turn off retries for LB worker | |48830|New|Nor|2010-03-01|IIS shutdown blocked in endpoint service when serv| |48891|Opn|Enh|2010-03-11|Missing EOL-style settings in tomcat/jk/trunk | |49035|New|Maj|2010-04-01|data lost when post a multipart/form-data form| |49063|New|Enh|2010-04-07|Please add JkStripSession status in jk-status work| |49135|New|Enh|2010-04-16|SPDY Connector for The Tomcat | |49469|New|Enh|2010-06-19|Workers status page has negative number of connect| |49732|Opn|Nor|2010-08-10|reply_timeout can't wait forever. | |49822|New|Enh|2010-08-25|Add hash lb worker method | |49903|New|Enh|2010-09-09|Make workers file reloadable | |50186|New|Nor|2010-10-31|Wrong documentation of connection_pool_timeout / c| |52334|New|Maj|2011-12-14|recover_time is not properly used | |52483|New|Enh|2012-01-18|Print JkOptions's options in log file and jkstatus| |52651|New|Nor|2012-02-13|JKSHMFile size limitation | |53324|Opn|Nor|2012-05-30|Starting with mod_jk 1.2.35 I cannot modify worker| |53542|New|Min|2012-07-13|Spelling mistake on 503 service unavailable page | |53762|New|Nor|2012-08-22|JK status manager: mass nodes handling doesn't wor| |53883|New|Maj|2012-09-17|isapi_redirect v 1.2.37 crashes w3wp.exe on the p| |53977|New|Maj|2012-10-07|32bits isapi connector cannot work in wow64 mode | |54027|New|Cri|2012-10-18|isapi send request to outside address instead of i| |54112|Opn|Blk|2012-11-07|ISAPI redirector not working when IIS recycles| |54117|New|Maj|2012-11-08|access violation exception in isapi_redirect.dll | |54177|New|Nor|2012-11-20|jkmanager generates non-well-formed XML for certai| |54596|New|Nor|2013-02-22|Relative path functionality truncates last charact| |54621|New|Nor|2013-02-28|[PATCH] custom mod_jk availability checks | |54646|New|Trv|2013-03-06|socket_keepalive is sometimes 1 or true or True in| |54923|New|Nor|2013-05-03|nsapi_redirect.so does not work with iPlanet on So|
Bug report for Tomcat 6 [2013/09/22]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=Critical REG=Regression MAJ=Major | | | | MIN=Minor NOR=NormalENH=Enhancement TRV=Trivial | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |41679|New|Enh|2007-02-22|SemaphoreValve should be able to filter on url pat| |41883|Ass|Enh|2007-03-18|use abstract wrapper instead of plain X509Certific| |43001|New|Enh|2007-07-30|JspC lacks setMappedFile and setDie for use in Ant| |43400|New|Enh|2007-09-14|enum support for tag libs | |43548|Opn|Enh|2007-10-04|xml schema for tomcat-users.xml | |43682|New|Enh|2007-10-23|JULI: web-inf/classes/logging.properties to suppor| |43742|New|Enh|2007-10-30|.tag compiles performed one at a time -- extremel| |43979|New|Enh|2007-11-27|Add abstraction for Java and Classfile output | |44199|New|Enh|2008-01-10|expose current backlog queue size | |44225|New|Enh|2008-01-14|SSL connector tries to load the private keystore f| |44294|New|Enh|2008-01-25|Support for EL functions with varargs | |44645|New|Enh|2008-03-20|[Patch] JNDIRealm - Doesn't support JNDI java.nam| |44787|New|Enh|2008-04-09|provide more error context on java.lang.IllegalSt| |45014|New|Enh|2008-05-15|Request and Response classes should have wrappers | |45282|New|Enh|2008-06-25|NioReceiver doesn't close cleanly, leaving sockets| |45832|New|Enh|2008-09-18|add DIGEST authentication support to Ant tasks| |45878|New|Enh|2008-09-24|Generated jars do not contain proper manifests or | |45879|Opn|Enh|2008-09-24|Windows installer fails to install NOTICE and RELE| |45931|Opn|Enh|2008-10-01|trimSpaces incorrectly modifies output| |46173|New|Enh|2008-11-09|Small patch for manager app: Setting an optional c| |46284|New|Enh|2008-11-24|Add flag to DeltaManager that blocks processing cl| |46350|New|Enh|2008-12-05|Maven repository should contain source bundles| |46558|Opn|Enh|2009-01-19|Shutdown port with address binding| |46902|New|Enh|2009-03-24|LoginValve to bypass restrictions of j_security_ch| |47214|New|Enh|2009-05-17|Inner classes that are explicitly referenced - sho| |47242|New|Enh|2009-05-22|request for AJP command line client | |47281|New|Enh|2009-05-28|Efficiency of the JDBCStore | |47407|New|Enh|2009-06-23|HttpSessionListener doesn't operate in the session| |47467|New|Enh|2009-07-02|Deployment of the war file by URL when contextpath| |47834|New|Enh|2009-09-14|TldConfig throws Exception when exploring unpacked| |47919|New|Enh|2009-09-30|Log Tomcat Java environment variables in additio| |48358|Opn|Enh|2009-12-09|JSP-unloading reloaded| |48543|New|Enh|2010-01-14|[Patch] More flexibility in specifying -Dcatalina.| |48672|New|Enh|2010-02-03|Tomcat Virtual Host Manager (/host-manager) have b| |48674|New|Enh|2010-02-03|Tomcat Virtual Host Manager application doesn't pe| |48743|New|Enh|2010-02-15|Make the SLEEP variable in catalina.sh settable fr| |48899|New|Enh|2010-03-12|Guess URI charset should solve lot of problems| |48922|New|Enh|2010-03-16|org.apache.catalina.connector.Request clone static| |48928|New|Enh|2010-03-17|An alternative solution to preloading classes when| |49161|New|Enh|2010-04-21|Unknown Publisher when installing tomcat 6.0.26 | |49176|Opn|Enh|2010-04-23|Jasper in Dev Mode Is Memory Inefficient | |49464|New|Enh|2010-06-18|DefaultServlet and CharacterEncoding | |49531|New|Enh|2010-06-30|singlesignon failover not working on DeltaManager/| |49804|New|Enh|2010-08-23|Allow Embedded.redirectStreams value to be configu| |49939|New|Enh|2010-09-16|Expose a method via JMX which empties the webapp f| |49943|New|Enh|2010-09-16|Logging (via juli) does not reread configuration c| |50285|New|Enh|2010-11-17|Standard HTTP and AJP connectors silently ignore a| |50288|New|Enh|2010-11-17|Uploading a war file that already exists should au| |50692|New|Enh|2011-01-31|Improve log message in ThreadPool.logFull | |51142|New|Enh|2011-05-03|Offer possible resolution of StringIndexOutOfBound| |51513|New|Enh|2011-07-15|GzipInterceptor: Do not compress small packages |
Bug report for Tomcat Native [2013/09/22]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=Critical REG=Regression MAJ=Major | | | | MIN=Minor NOR=NormalENH=Enhancement TRV=Trivial | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |46179|Opn|Maj|2008-11-10|apr ssl client authentication | |48655|Inf|Nor|2010-02-02|Active multipart downloads prevent tomcat shutdown| |49038|Inf|Nor|2010-04-02|Crash in tcnative | |52153|New|Maj|2011-11-08|periodic JVM crash (access violation) on buffer fl| |52319|New|Maj|2011-12-12|Tomcat 6 crashes with [libapr-1.so.0+0x196da] sig| |52627|New|Min|2012-02-08|Segmentation fault in org.apache.tomcat.jni.File.i| |53605|New|Nor|2012-07-26|use tcnative-1.1.24 Tomcat shutdown still crash | |53847|Inf|Nor|2012-09-10|High CPU usage in tomcat native 1.22+ | |53940|New|Enh|2012-09-27|Added support for new CRL loading after expiration| |53952|New|Nor|2012-10-02|Add support for TLS 1.1 and 1.2 | |54085|New|Nor|2012-11-01|ssl_socket_recv sometimes loops infinitely with no| |54664|New|Reg|2013-03-11|[1.1.27 branch] Poll.remove incorrectly reports AP| |55087|New|Cri|2013-06-10|tomcat crashes in tcnative-1.dll with OCSP when OC| |55113|New|Nor|2013-06-18|FIPS-compatible OpenSSL fails fingerprint test in | |55114|New|Nor|2013-06-18|BUILDING file in win32 source package contains UNI| +-+---+---+--+--+ | Total 15 bugs | +---+ - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Bug report for Tomcat Modules [2013/09/22]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=Critical REG=Regression MAJ=Major | | | | MIN=Minor NOR=NormalENH=Enhancement TRV=Trivial | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |48240|New|Nor|2009-11-19|Tomcat-Lite missing @Override markers | |48268|New|Nor|2009-11-23|Patch to fix generics in tomcat-lite | |48861|New|Nor|2010-03-04|Files without AL headers | |49685|New|Nor|2010-08-02|Unsafe synchronization in class ManagedBean | |49686|New|Nor|2010-08-02|Using an instance lock to protect static shared da| |50571|Inf|Nor|2011-01-11|Tomcat 7 JDBC connection pool exception enhancemen| |51595|Inf|Nor|2011-08-01|org.apache.tomcat.jdbc.pool.jmx.ConnectionPool sho| |51879|Inf|Enh|2011-09-22|Improve access to Native Connection Methods | |52024|Inf|Enh|2011-10-13|Custom interceptor to support automatic failover o| |53088|Opn|Min|2012-04-17|Give PoolCleaner TimerTask a better name | |53198|New|Cri|2012-05-07|'driverClassName' Data Source Property Being Manda| |53199|Inf|Enh|2012-05-07|Refactor ConnectionPool to use ScheduledExecutorSe| |53200|New|Enh|2012-05-07|Be able to use SlowQueryReport without reporting f| |53770|New|Enh|2012-08-23|tomcat-pool: always log validation query syntax er| |53853|New|Nor|2012-09-11|Can tomcat-jdbc consider Thread#getContextClassLoa| |54225|New|Nor|2012-11-30|if initSQL property is set to an empty string a Nu| |54227|New|Nor|2012-11-30|maxAge should be checked on borrow| |54235|New|Nor|2012-12-03|tomcat jdbc pool stackoverflow error used with spr| |54395|New|Nor|2013-01-09|JdbcInterceptor config parameter parsing errors | |54437|New|Enh|2013-01-16|Update PoolProperties javadoc for ConnectState int| |54537|New|Cri|2013-02-07|StatementFinalizer closeInvoked is too slow for la| |54929|New|Nor|2013-05-05|jdbc-pool cannot be used with Java 1.5, java.lang| |54978|New|Nor|2013-05-15|Validate on Borrow should be tested on Reconnect i| |55078|New|Nor|2013-06-07|Configuring a DataSource Resource with dataSourceJ| |55444|New|Nor|2013-08-18|Support JDBC Drivers outside of tomcat/lib/ | +-+---+---+--+--+ | Total 25 bugs | +---+ - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Bug report for Tomcat 7 [2013/09/22]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=Critical REG=Regression MAJ=Major | | | | MIN=Minor NOR=NormalENH=Enhancement TRV=Trivial | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |18500|New|Enh|2003-03-30|Host aliases to match by regular expression | |28039|Opn|Enh|2004-03-30|Cluster Support for SingleSignOn | |40881|Opn|Enh|2006-11-02|Unable to receive message through TCP channel - | |41007|Opn|Enh|2006-11-20|Can't define customized 503 error page| |43866|New|Enh|2007-11-14|add support for session attribute propagation with| |43925|Opn|Enh|2007-11-21|org.apache.jasper.runtime.BodyContentImpl causing | |44216|New|Enh|2008-01-11|Don't reuse session ID even if emptySessionPath=tr| |49395|New|Enh|2010-06-06|manager.findLeaks : display the date when the leak| |49589|New|Enh|2010-07-12|Tag handlers with constant attribute values are al| |49785|New|Enh|2010-08-19|Enabling TLS for JNDIRealm| |49821|New|Enh|2010-08-25|Tomcat CLI| |50019|New|Enh|2010-09-28|Adding JNDI lookup-name support In XML and Resou| |50175|New|Enh|2010-10-28|Enhance memory leak detection by selectively apply| |50234|New|Enh|2010-11-08|JspC use servlet 3.0 features | |50504|New|Enh|2010-12-21|Allow setting query string character set trough re| |50670|New|Enh|2011-01-27|Tribes | RpcChannel | Add option to specify extern| |51195|New|Enh|2011-05-13|Find leaks reports a false positive memory/class| |51294|Opn|Enh|2011-05-30|Since 7.0.12 do not work option unpackWARs=true fo| |51423|Inf|Enh|2011-06-23|[Patch] to add a path and a version parameters to | |51463|New|Enh|2011-07-01|Tomcat.setBaseDir (package org.apache.catalina.st| |51496|New|Enh|2011-07-11|NSIS - Warn that duplicate service name will resul| |51587|New|Enh|2011-07-29|Implement status and uptime commands | |51953|New|Enh|2011-10-04|Proposal: netmask filtering valve and filter | |52235|New|Enh|2011-11-23|Please do a bit of SEO tuning for the web site| |52323|New|Enh|2011-12-13|Cobertura test code coverage support for build.xml| |52381|New|Enh|2011-12-22|Please add OSGi metadata | |52448|New|Enh|2012-01-11|Cache jar indexes in WebappClassLoader to speed up| |52489|New|Enh|2012-01-19|Enhancement request for code signing of war files | |52688|New|Enh|2012-02-16|Add ability to remove old access log files| |52751|Opn|Enh|2012-02-23|Optimized configuration of the system info display| |52952|New|Enh|2012-03-20|Improve ExtensionValidator handling for embedded s| |53085|New|Enh|2012-04-16|[perf] [concurrency] DefaultInstanceManager.annota| |53387|New|Enh|2012-06-08|SSI: Allow to use $1 to get result of regular expr| |53411|Opn|Enh|2012-06-13|NullPointerException in org.apache.tomcat.util.buf| |53492|New|Enh|2012-07-01|Make JspC shell multithreaded | |53553|New|Enh|2012-07-16|[PATCH] Deploy uploaded WAR with context.xml from | |53620|New|Enh|2012-07-30|[juli] delay opening a file until something gets l| |54330|New|Enh|2012-12-19|Patch with some refactoring of Member.java| |54499|New|Enh|2013-01-29|Implementation of Extensible EL Interpreter | |54618|New|Enh|2013-02-28|Add filter implementing HTTP Strict Transport Secu| |54802|New|Enh|2013-04-04|Provide location information for exceptions thrown| |55104|New|Enh|2013-06-16|Allow passing arguments with spaces to Commons Dae| |55317|New|Enh|2013-07-28|Facilitate weaving by allowing ClassFileTransforme| |55470|New|Enh|2013-08-23|Help users for ClassNotFoundExceptions during star| |55477|New|Enh|2013-08-23|Add a solution to map an realm name to a security | |55530|New|Enh|2013-09-06|webdavServlet support aliases folder upload and de| |55576|New|Maj|2013-09-20|Order of ServletRequest parameters is not preserve| +-+---+---+--+--+ | Total 47 bugs | +---+ - To unsubscribe,
Re: SCI discovery ordering
On 22/09/2013 00:27, Jeremy Boynes wrote: If the intent of the language in the spec is to give the application implementation priority if it comes first in the delegation order, this should apply to discovery as well as classloading per the ServiceLoader semantic. No, the intent of the language was only to make clear that the delegation order must be honoured when loading SCI implementations. To quote from the Servlet EG discussion on this topic: quote If the same ServletContainerInitializer is both on the container classpath and in a jar in WEB-INF/lib, which one should be used? Will this be by the normal inverted classloading priority of web applications? /quote The result of the discussion was Yes. and the language was added to the end of 8.2. It went on to discuss some edge cases but no conclusion was reached and no further clarification was added to the spec. r1524727's scheme is more nuanced: * if orderedLibs is present, it loads and executes in the following order: 1) SCIs defined in the webapp's parent classloader (container SCIs) in undefined order 2) SCIs defined by jars in orderedLibs (application SCIs) in orderedLibs order * else, if the webapp loader is application first: 1) SCIs defined by the application classloader in undefined order 2) SCIs defined by the webapp's parent classloader in undefined order * else, as the webapp loader is parent first: 1) SCIs defined by the webapp's parent classloader in undefined order 2) SCIs defined by the application classloader in undefined order I'll take another look at r1524727 as that isn't want was intended. What should happen is: - List all SCIs defined by the container - List all SCIs defined by the application (may only be from a subset of JARs if absolute ordering is present) - Load all the SCIs defined by the container (note some may be loaded from the application if the application ships with an alternative implementation and the application first delegation model is being used) - Load all the SCIs defined by the application As a concrete example of how this impacts the behaviour, consider the case where the application includes its own JSP engine. With the RI's delegation model, the application's engine's SCI would execute first allowing it to register a Servlet and mapping to handle the *.jsp pattern. When the container's engine's SCI was executed, that mapping would already be bound and could not be pre-empted (engines already need to allow for that in case the application configured that mapping in its web.xml). If the container is always given priority, then the container's engine would be used rather than the one the application intended. No. It would still be loaded from the application (for that webapp). The spec is very clear that SCIs are intended for frameworks and not for general application initialization (which is covered by context listeners and fragment ordering). I disagree. The specification uses adding frameworks as an example of a use case the an SCI can support, but nowhere does it say that it may not be used by applications or that it isn't intended for applications. Mark - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
[Bug 55582] New: Concurrent issue of TagFileProcessor
https://issues.apache.org/bugzilla/show_bug.cgi?id=55582
Bug ID: 55582
Summary: Concurrent issue of TagFileProcessor
Product: Tomcat 7
Version: trunk
Hardware: All
OS: All
Status: NEW
Severity: normal
Priority: P2
Component: Jasper
Assignee: dev@tomcat.apache.org
Reporter: xs...@ebay.com
Created attachment 30871
-- https://issues.apache.org/bugzilla/attachment.cgi?id=30871action=edit
Unsynchronized getting wrapper from RuntimeContext
The following code has concurrent issue.
JspRuntimeContext rctxt = ctxt.getRuntimeContext();
JspServletWrapper wrapper = rctxt.getWrapper(wrapperUri);
synchronized (rctxt) {
if (wrapper == null) {
It creates duplicated JspServletWrapper in this scenario,
A.jsp -- C.tag
B.jsp -- C.tag
A.jsp and B.jsp are both compiling and come to the given lines.
Two threads all get null from JspRuntimeContext(JspServletWrapper == null).
So two instances of JspServletWrapper was created.
--
You are receiving this mail because:
You are the assignee for the bug.
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
[Bug 55582] Concurrent issue of TagFileProcessor
https://issues.apache.org/bugzilla/show_bug.cgi?id=55582
Sheldon Shao xs...@ebay.com changed:
What|Removed |Added
Attachment #30871|0 |1
is patch||
--- Comment #1 from Sheldon Shao xs...@ebay.com ---
Comment on attachment 30871
-- https://issues.apache.org/bugzilla/attachment.cgi?id=30871
Unsynchronized getting wrapper from RuntimeContext
Index: java/org/apache/jasper/compiler/TagFileProcessor.java
===
--- java/org/apache/jasper/compiler/TagFileProcessor.java (revision
1507186)
+++ java/org/apache/jasper/compiler/TagFileProcessor.java (working copy)
@@ -533,9 +533,11 @@
JspCompilationContext ctxt = compiler.getCompilationContext();
JspRuntimeContext rctxt = ctxt.getRuntimeContext();
-JspServletWrapper wrapper = rctxt.getWrapper(wrapperUri);
+
synchronized (rctxt) {
+ JspServletWrapper wrapper = rctxt.getWrapper(wrapperUri);
+
if (wrapper == null) {
wrapper = new JspServletWrapper(ctxt.getServletContext(), ctxt
.getOptions(), tagFilePath, tagInfo, ctxt
--
You are receiving this mail because:
You are the assignee for the bug.
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
Issue in AprEndpoint detected by TestCoyoteAdapter
I debugged around my occasional failures for TestCoyoteAdapter when using APR. Error is: SEVERE [http-apr-127.0.0.1-auto-13-Poller] org.apache.tomcat.util.net.AprEndpoint$Poller.run Poller failed with error [81] : [File descriptor in bad state] In the native world it is: 19372/128: port_associate(6, 4, 0x, 0x0001, 0x0089ABE8) Err#81 EBADFD The bad FD here is 0x. Normally the FDs put into this slot of port_associate during the test are small numbers between 6 and 20 (decimal). Here's a -1 and we must have an error before. Wat we would have expected is FD 11. In the truss snippet, Thread 2 is the client opening the connection, thread 130 accepts it and thread 133 starts to work on it: 19372/2:so_socket(PF_INET, SOCK_STREAM, IPPROTO_IP, , SOV_DEFAULT) = 8 Connect 19372/2:connect(8, 0xFDF7DAE0, 16, SOV_DEFAULT) = 0 19372/2:AF_INET name = 127.0.0.1 port = 42932 Accept 19372/130: accept(5, 0xB397F608, 0xB397F5F4, SOV_DEFAULT) = 11 19372/130: AF_INET name = 127.0.0.1 port = 42936 19372/2:getsockname(8, 0xFDF7DAE0, 0xFDF7DA2C, SOV_DEFAULT) = 0 19372/2:AF_INET name = 127.0.0.1 port = 42936 19372/130: fcntl(11, F_GETFD, 0x0003) = 0 19372/130: fcntl(11, F_SETFD, 0x0001) = 0 19372/2:setsockopt(8, tcp, TCP_NODELAY, 0xFDF7DE10, 4, SOV_DEFAULT) = 0 19372/130: lwp_cond_signal(0x0031E3B0) = 0 19372/130: condvar type: USYNC_THREAD Send request 19372/2:send(8, G E T / f o o ; j s e.., 227, 0) = 227 19372/13: lwp_cond_wait(0x0031E3B0, 0x0031E398, 0xB457F278, 0) = 0 19372/13: condvar type: USYNC_THREAD 19372/13: mutex type: USYNC_THREAD 19372/13: timeout: 0.98773 sec 19372/130: lwp_cond_signal(0x004215F0) = 0 19372/130: condvar type: USYNC_THREAD 19372/133: lwp_cond_wait(0x004215F0, 0x004215D8, 0x, 0) = 0 19372/133: condvar type: USYNC_THREAD 19372/133: mutex type: USYNC_THREAD 19372/133: setsockopt(11, tcp, TCP_NODELAY, 0xB3C7F40C, 4, SOV_DEFAULT) = 0 19372/133: fcntl(11, F_GETFL) = 2 FWRITE 19372/133: fcntl(11, F_SETFL, FWRITE|FNONBLOCK)= 0 19372/133: lwp_cond_signal(0x0031E3B0) = 0 19372/133: condvar type: USYNC_THREAD 19372/133: lwp_cond_signal(0x00274448) = 0 19372/133: condvar type: USYNC_THREAD 19372/128: lwp_cond_wait(0x00274448, 0x00274430, 0xB417F580, 0) = 0 19372/128: condvar type: USYNC_THREAD 19372/128: mutex type: USYNC_THREAD 19372/128: timeout: 9.97356 sec 19372/128: lwp_cond_signal(0x0031E3B0) = 0 19372/128: condvar type: USYNC_THREAD 19372/128: lwp_cond_signal(0x0031E3B0) = 0 19372/128: condvar type: USYNC_THREAD Adding to the pollset with wrong FD -1 instead of 11. 19372/128: port_associate(6, 4, 0x, 0x0001, 0x0089ABE8) Err#81 EBADFD I added some Java log statements to track calls to the Poll class (please ignore that hey are log level warn): 22-Sep-2013 12:19:12.909 FINE [http-apr-127.0.0.1-auto-13-Poller] org.apache.tomcat.util.net.AprEndpoint$Poller.run socket [2,218,784] 22-Sep-2013 12:19:12.910 WARNING [http-apr-127.0.0.1-auto-13-Poller] org.apache.tomcat.util.net.AprEndpoint$Poller.addToPoller Adding to poller number 0 of size 0 socket 2218784 22-Sep-2013 12:19:12.911 WARNING [http-apr-127.0.0.1-auto-13-Poller] org.apache.tomcat.util.net.AprEndpoint$Poller.addToPoller Adding to poller number 0 returned with 0 Here begins the root cause: - start polling poller of size 1, containing only socket 2218784: 22-Sep-2013 12:19:12.912 WARNING [http-apr-127.0.0.1-auto-13-Poller] org.apache.tomcat.util.net.AprEndpoint$Poller.run Polling poller number 0 of size 1 with timeout 2000 - start removing from the same poller concurrently in the exec thread: 22-Sep-2013 12:19:12.914 WARNING [http-apr-127.0.0.1-auto-13-exec-10] org.apache.tomcat.util.net.AprEndpoint$Poller.removeFromPoller Removing from poller number 0 of size 1 socket 2856392 - polling returns with one fd and decreases size to 0: 22-Sep-2013 12:19:12.913 WARNING [http-apr-127.0.0.1-auto-13-Poller] org.apache.tomcat.util.net.AprEndpoint$Poller.run Polling poller number 0 returned with 1 - remove returns with success and decreases size to -1! : 22-Sep-2013 12:19:12.915 WARNING [http-apr-127.0.0.1-auto-13-exec-10] org.apache.tomcat.util.net.AprEndpoint$Poller.removeFromPoller Removing from poller number 0 returned with 0 22-Sep-2013 12:19:12.916 FINE [http-apr-127.0.0.1-auto-13-Poller] org.apache.tomcat.util.net.AprEndpoint$Poller.run Processing socket [2,218,784] for event(s) [1] 22-Sep-2013
Re: Issue in AprEndpoint detected by TestCoyoteAdapter
On 22.09.2013 13:17, Rainer Jung wrote:
I debugged around my occasional failures for TestCoyoteAdapter when
using APR.
Error is:
SEVERE [http-apr-127.0.0.1-auto-13-Poller]
org.apache.tomcat.util.net.AprEndpoint$Poller.run Poller failed with
error [81] : [File descriptor in bad state]
...
Not sure whether the problem is more in the concurrent poll plus remove,
or the following code in poll.c:
TCN_IMPLEMENT_CALL(jint, Poll, remove)(TCN_STDARGS, jlong pollset,
jlong socket)
{
apr_pollfd_t fd;
apr_status_t rv;
tcn_pollset_t *p = J2P(pollset, tcn_pollset_t *);
tcn_socket_t *s = J2P(socket, tcn_socket_t *);
UNREFERENCED_STDARGS;
TCN_ASSERT(socket != 0);
if (s-pe == NULL) {
/* Already removed */
return APR_SUCCESS;
}
Here we return APR_SUCCESS and the code calling Poll.remove in
AprEndpoint always does:
rv = Poll.remove(pollers[i], socket);
if (rv != Status.APR_NOTFOUND) {
pollerSpace[i]++;
connectionCount--;
break;
}
So the pollerSpace and connectionCount numbers are (in/de)cremented.
The following patch seems to fix it for me, at least 150 test runs for
TestCoyoteAdapter were successful:
Index: ../native/branches/1.1.x/native/src/poll.c
===
--- ../native/branches/1.1.x/native/src/poll.c (revision 1525348)
+++ ../native/branches/1.1.x/native/src/poll.c (working copy)
@@ -259,7 +259,7 @@
if (s-pe == NULL) {
/* Already removed */
-return APR_SUCCESS;
+return APR_NOTFOUND;
}
fd.desc_type = APR_POLL_SOCKET;
fd.desc.s = s-sock;
I'll let the full test suite for APR run a couple of times as a more
thorough check. I hope the change doesn't break the correctnes of the
connection counting which is only done after succesful return of remove.
Regards,
Rainer
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: Issue in AprEndpoint detected by TestCoyoteAdapter
On 22.09.2013 15:39, Rainer Jung wrote:
On 22.09.2013 13:17, Rainer Jung wrote:
I debugged around my occasional failures for TestCoyoteAdapter when
using APR.
Error is:
SEVERE [http-apr-127.0.0.1-auto-13-Poller]
org.apache.tomcat.util.net.AprEndpoint$Poller.run Poller failed with
error [81] : [File descriptor in bad state]
...
The following patch seems to fix it for me, at least 150 test runs for
TestCoyoteAdapter were successful:
Index: ../native/branches/1.1.x/native/src/poll.c
===
--- ../native/branches/1.1.x/native/src/poll.c (revision 1525348)
+++ ../native/branches/1.1.x/native/src/poll.c (working copy)
@@ -259,7 +259,7 @@
if (s-pe == NULL) {
/* Already removed */
-return APR_SUCCESS;
+return APR_NOTFOUND;
}
fd.desc_type = APR_POLL_SOCKET;
fd.desc.s = s-sock;
I'll let the full test suite for APR run a couple of times as a more
thorough check. I hope the change doesn't break the correctnes of the
connection counting which is only done after succesful return of remove.
I see only failures in TestCoyoteOutputStream when running the full test
suite for APR, all other tests look OK.
When running TestCoyoteOutputStream alone in a loop it also occasionally
(2/10) fails for bio and nio, for apr somewhat more often (4/10), so the
failure seems unrelated to the poll.c change. Actually there's two types
of failures for TestCoyoteOutputStream, mostly a 300 seconds read
timeout in testNonBlockingWriteNoneBlockingWriteOnce and once for nio
java.io.IOException: Bogus chunk size in
testNonBlockingWriteOnceBlockingWriteOnce
Regards,
Rainer
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: SCI discovery ordering
On Sep 22, 2013, at 1:44 AM, Mark Thomas ma...@apache.org wrote: On 22/09/2013 00:27, Jeremy Boynes wrote: As a concrete example of how this impacts the behaviour, consider the case where the application includes its own JSP engine. With the RI's delegation model, the application's engine's SCI would execute first allowing it to register a Servlet and mapping to handle the *.jsp pattern. When the container's engine's SCI was executed, that mapping would already be bound and could not be pre-empted (engines already need to allow for that in case the application configured that mapping in its web.xml). If the container is always given priority, then the container's engine would be used rather than the one the application intended. No. It would still be loaded from the application (for that webapp). For two versions of the *same* implementation, yes. But if they used *different* implementations of the *same functionality,* the container's would always get precedence. For example, if an application included Tyrus's WebSocket implementation it would always be invoked after Tomcat's. Or for JAX-RS, if the container was configured with CXF and the application included Jersey, Jersey would not be able to register its Servlets for the REST endpoints as CXF would have already mapped them. The issue here is that programmatic registrations cannot modify the existing configuration. Once a framework has registered a servlet, filter, listener or mapping it cannot be replaced by another. Frameworks that applications bundle in WEB-INF/lib need to have a chance to perform their initialization before an equivalent but different implementation provided by the container. -- Jeremy signature.asc Description: Message signed with OpenPGP using GPGMail
Re: SCI discovery ordering
On Sep 22, 2013, at 10:55 AM, Jeremy Boynes jboy...@apache.org wrote: On Sep 22, 2013, at 1:44 AM, Mark Thomas ma...@apache.org wrote: On 22/09/2013 00:27, Jeremy Boynes wrote: As a concrete example of how this impacts the behaviour, consider the case where the application includes its own JSP engine. With the RI's delegation model, the application's engine's SCI would execute first allowing it to register a Servlet and mapping to handle the *.jsp pattern. When the container's engine's SCI was executed, that mapping would already be bound and could not be pre-empted (engines already need to allow for that in case the application configured that mapping in its web.xml). If the container is always given priority, then the container's engine would be used rather than the one the application intended. No. It would still be loaded from the application (for that webapp). For two versions of the *same* implementation, yes. But if they used *different* implementations of the *same functionality,* the container's would always get precedence. For example, if an application included Tyrus's WebSocket implementation it would always be invoked after Tomcat's. Or for JAX-RS, if the container was configured with CXF and the application included Jersey, Jersey would not be able to register its Servlets for the REST endpoints as CXF would have already mapped them. The issue here is that programmatic registrations cannot modify the existing configuration. Once a framework has registered a servlet, filter, listener or mapping it cannot be replaced by another. Frameworks that applications bundle in WEB-INF/lib need to have a chance to perform their initialization before an equivalent but different implementation provided by the container. Patch to avoid this by following classloader delegation order … sci.patch Description: Binary data signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [VOTE] Release Apache Tomcat 7.0.43
2013/9/21 Mark Thomas ma...@apache.org On 20/09/2013 08:38, Violeta Georgieva wrote: The proposed Apache Tomcat 7.0.43 release is now available for voting. This release candidate contains JSR-356 Java WebSocket 1.0 implementation. Note that use of this functionality requires Java 7. It can be obtained from: https://dist.apache.org/repos/dist/dev/tomcat/tomcat-7/v7.0.43/ The Maven staging repo is: https://repository.apache.org/content/repositories/orgapachetomcat-082/ The svn tag is: http://svn.apache.org/repos/asf/tomcat/tc7.0.x/tags/TOMCAT_7_0_43/ The proposed 7.0.43 release is: [X] Broken - do not release I'm going to mark 7.0.43 as not released. Tagging 7.0.44 in the afternoon. Regards Violeta
svn commit: r2947 - /dev/tomcat/tomcat-7/v7.0.43/
Author: violetagg Date: Mon Sep 23 04:19:29 2013 New Revision: 2947 Log: 7.0.43 RC did not pass the vote - dropping. Removed: dev/tomcat/tomcat-7/v7.0.43/ - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
svn commit: r2947 - /dev/tomcat/tomcat-7/v7.0.43/
Author: violetagg Date: Mon Sep 23 04:19:29 2013 New Revision: 2947 Log: 7.0.43 RC did not pass the vote - dropping. Removed: dev/tomcat/tomcat-7/v7.0.43/ - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
svn commit: r1525505 - /tomcat/tc7.0.x/trunk/webapps/docs/changelog.xml
Author: violetagg Date: Mon Sep 23 04:42:31 2013 New Revision: 1525505 URL: http://svn.apache.org/r1525505 Log: Mark 7.0.43 as not released based on the vote results. Modified: tomcat/tc7.0.x/trunk/webapps/docs/changelog.xml Modified: tomcat/tc7.0.x/trunk/webapps/docs/changelog.xml URL: http://svn.apache.org/viewvc/tomcat/tc7.0.x/trunk/webapps/docs/changelog.xml?rev=1525505r1=1525504r2=1525505view=diff == --- tomcat/tc7.0.x/trunk/webapps/docs/changelog.xml (original) +++ tomcat/tc7.0.x/trunk/webapps/docs/changelog.xml Mon Sep 23 04:42:31 2013 @@ -55,7 +55,7 @@ They eventually become mixed with the numbered issues. (I.e., numbered issues to not pop up wrt. others). -- -section name=Tomcat 7.0.43 (voiletagg) +section name=Tomcat 7.0.43 (violetagg) rtext=not released subsection name=Catalina changelog add - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: Issue in AprEndpoint detected by TestCoyoteAdapter
On 09/22/2013 03:39 PM, Rainer Jung wrote:
On 22.09.2013 13:17, Rainer Jung wrote:
I debugged around my occasional failures for TestCoyoteAdapter when
using APR.
Error is:
SEVERE [http-apr-127.0.0.1-auto-13-Poller]
org.apache.tomcat.util.net.AprEndpoint$Poller.run Poller failed with
error [81] : [File descriptor in bad state]
...
Not sure whether the problem is more in the concurrent poll plus remove,
or the following code in poll.c:
TCN_IMPLEMENT_CALL(jint, Poll, remove)(TCN_STDARGS, jlong pollset,
jlong socket)
{
apr_pollfd_t fd;
apr_status_t rv;
tcn_pollset_t *p = J2P(pollset, tcn_pollset_t *);
tcn_socket_t *s = J2P(socket, tcn_socket_t *);
UNREFERENCED_STDARGS;
TCN_ASSERT(socket != 0);
if (s-pe == NULL) {
/* Already removed */
return APR_SUCCESS;
}
Here we return APR_SUCCESS and the code calling Poll.remove in
AprEndpoint always does:
rv = Poll.remove(pollers[i], socket);
if (rv != Status.APR_NOTFOUND) {
pollerSpace[i]++;
connectionCount--;
break;
}
So the pollerSpace and connectionCount numbers are (in/de)cremented.
The following patch seems to fix it for me, at least 150 test runs for
TestCoyoteAdapter were successful:
Index: ../native/branches/1.1.x/native/src/poll.c
===
--- ../native/branches/1.1.x/native/src/poll.c (revision 1525348)
+++ ../native/branches/1.1.x/native/src/poll.c (working copy)
@@ -259,7 +259,7 @@
if (s-pe == NULL) {
/* Already removed */
-return APR_SUCCESS;
+return APR_NOTFOUND;
}
fd.desc_type = APR_POLL_SOCKET;
fd.desc.s = s-sock;
The patch seems fine. I mean any return value should do in theory.
The main question is why is particular socket removed twice
from the Poller. This is called directly from java code
so wrapper seems to call it twice (or more).
I suspect that the socket is first closed and then Poller loop removes it.
Or it can be removed by poll with doRemove == true or during pollset maintain.
In any case after removed either by poll or maintain returned set of removed
sockets must be invalidated from pollset so it doesn't use it again in explicit
remove.
Regards
--
^TM
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
