[
https://issues.apache.org/jira/browse/LOG4J2-1028?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14597324#comment-14597324
]
Gary Gregory edited comment on LOG4J2-1028 at 6/23/15 8:17 AM:
---------------------------------------------------------------
Stefan,
The new patch applies cleanly, thank you for the refresh.
Unfortunately, as I mentioned previously \[1], the patch causes some tests to
fail. Each failing test should be reviewed carefully to determine if the test
should be changed or if the patch is really valid.
In the Core module:
{noformat}
Failed tests:
AsyncAppenderTest.rewriteTest:75 Expected
org.apache.logging.log4j.core.appender.AsyncAppenderTest rewriteTest This is a
test, Actual org.apache.logging.log4j.spi.AbstractLogger logMessage This is a
test
MemoryMappedFileAppenderLocationTest.testMemMapLocation:75 Shrunk to actual
used size expected:<478> but was:<356>
AsyncLoggerConfigTest.testAdditivity:62 location
AsyncLoggerLocationTest.testAsyncLogWritesToLog:68 has location
Rfc5424LayoutTest.testLoggerFields:336 No class/method
Rfc5424LayoutTest.testMDCLoggerFields:289 No class/method
CallerInformationTest.testClassLogger:44 Incorrect caller class name.
expected:<...pache.logging.log4j.[core.pattern.CallerInformationTest]> but
was:<...pache.logging.log4j.[spi.AbstractLogger]>
CallerInformationTest.testMethodLogger:60 Incorrect caller method name.
expected:<[testMethodLogger]> but was:<[logMessage]>
NoConsoleNoAnsiTest.testReplacement:58 Replacement failed - expected ending
ERROR LoggerTest o.a.l.l.c.p.NoConsoleNoAnsiTest
org.apache.logging.log4j.core.pattern.NoConsoleNoAnsiTest
, actual 2015-06-23 01:12:25,330 ERROR LoggerTest o.a.l.l.s.AbstractLogger
org.apache.logging.log4j.core.pattern.NoConsoleNoAnsiTest
StyleConverterTest.testReplacement:59 Replacement failed - expected ending
←[1;31mERROR←[m ←[1;36mLoggerTest←[m o.a.l.l.c.p.StyleConverterTest
org.apache.logging.log4j.core.pattern.StyleConverterTes
t
, actual 2015-06-23 01:12:31,475 ←[1;31mERROR←[m ←[1;36mLoggerTest←[m
o.a.l.l.s.AbstractLogger
org.apache.logging.log4j.core.pattern.StyleConverterTest
Tests run: 764, Failures: 10, Errors: 0, Skipped: 29
{noformat}
Thank you,
Gary
\[1]
https://issues.apache.org/jira/browse/LOG4J2-1028?focusedCommentId=14581253&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14581253
was (Author: garydgregory):
Stefan,
The new patch applies cleanly, thank you for the refresh.
Unfortunately, as I mentioned previously [1], the patch causes some tests to
fail. Each failing test should be reviewed carefully to determine if the test
should be changed or if the patch is really valid.
Thank you,
Gary
[1]
https://issues.apache.org/jira/browse/LOG4J2-1028?focusedCommentId=14581253&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14581253
> Incorrect Line Number in location information
> ---------------------------------------------
>
> Key: LOG4J2-1028
> URL: https://issues.apache.org/jira/browse/LOG4J2-1028
> Project: Log4j 2
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.3
> Environment: Java 1.8.0_45 64-bit
> Maven 3.3.1
> Windows 7
> basically any...
> Reporter: Stefan Leonhartsberger
> Attachments: logLinePatch.diff
>
>
> When enabling location information in Log4j2
> The line numbers of the log records are incorrect
> as they are only taking into account the first time in the stacktrace where
> the
> class in question (FQCN) was called
> This however may not be accurate as the class may call methods within itself
> adding to the stacktrace - these are not considered when gathering location
> information within Log4jLogEvent.
> An example stacktrace:
> {code}
> stackTrace = {java.lang.StackTraceElement[18]@4612}
> 0 = {java.lang.StackTraceElement@4617}
> "java.lang.Thread.getStackTrace(Thread.java:1552)"
> 1 = {java.lang.StackTraceElement@4618}
> "org.apache.logging.log4j.core.impl.Log4jLogEvent.calcLocation(Log4jLogEvent.java:386)"
> 2 = {java.lang.StackTraceElement@4619}
> "org.apache.logging.log4j.core.async.AsyncLogger.location(AsyncLogger.java:293)"
> 3 = {java.lang.StackTraceElement@4620}
> "org.apache.logging.log4j.core.async.AsyncLogger.logMessage(AsyncLogger.java:252)"
> 4 = {java.lang.StackTraceElement@4621}
> "org.apache.logging.log4j.spi.AbstractLogger.logMessage(AbstractLogger.java:738)"
> 5 = {java.lang.StackTraceElement@4622}
> "org.apache.logging.log4j.spi.AbstractLogger.logIfEnabled(AbstractLogger.java:708)"
> 6 = {java.lang.StackTraceElement@4623}
> "com.test.logging.factory.Log4j2Factory$Log4J2CatsLogger.log(Log4j2Factory.java:106)"
> 7 = {java.lang.StackTraceElement@4624}
> "com.test.agent.configuration.Configuration.setPttId(Configuration.java:1079)"
> 8 = {java.lang.StackTraceElement@4625}
> "com.test.agent.configuration.Configuration.loadDefaultConfiguration(Configuration.java:323)"
> 9 = {java.lang.StackTraceElement@4614}
> "com.test.agent.configuration.Configuration.<init>(Configuration.java:302)"
> 10 = {java.lang.StackTraceElement@4613}
> "com.test.agent.configuration.ConfigurationParser.createConfig(ConfigurationParser.java:51)"
> 11 = {java.lang.StackTraceElement@4626}
> "com.test.agent.AgentMain.initConfig(AgentMain.java:231)"
> 12 = {java.lang.StackTraceElement@4627}
> "com.test.agent.AgentMain.init(AgentMain.java:204)"
> 13 = {java.lang.StackTraceElement@4628}
> "com.c4i.test.agent.AgentBase.init(AgentBase.java:456)"
> 14 = {java.lang.StackTraceElement@4629}
> "com.test.agent.AgentBase.createAgent(AgentBase.java:194)"
> 15 = {java.lang.StackTraceElement@4630}
> "com.test.agent.AgentBase.agentmain(AgentBase.java:141)"
> 16 = {java.lang.StackTraceElement@4631}
> "com.test.agent.AgentMain.main(AgentMain.java:109)"
> 17 = {java.lang.StackTraceElement@4632}
> "com.test.agent.AgentMain.main(AgentMain.java:86)"
> {code}
> Expected Line Number: 1079
> Received Line Number: 302
> Reason is that the stack is iterated from the wrong direction - i=17.
> (starting with i=0 would result in the correct behaviour.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
