Skip to main content
aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2020-06-10Merge branch osgiR8 into masterI20200612-0400I20200611-1800osgiR8Thomas Watson42-1035/+388
2020-06-10Update OSGi R8 API to the latestThomas Watson4-36/+45
2020-06-08Bug 564065 - Add org.junit.jupiter.api bundleI20200610-1800I20200609-1800I20200609-0150I20200608-1800Thomas Watson1-0/+5
Change-Id: Ia23d93226ef0dc810dd9404f1b7d4fc460560a88 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-06-08Update versions for 4.17Thomas Watson10-10/+10
Change-Id: Iee39638253ae4479e745dc71fe173c3dd080df73 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-06-08Binaries v20200608-1138Eclipse Releng Bot6-6/+6
2020-06-07Bug 563845 - Version bumps for 4.17 streamI20200607-1800Alexander Kurtakov2-2/+2
Change-Id: Iea169b4d5eff2c1e1264bab5a526477ec0bb3f58 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2020-06-04Bug 563840 - POM and product version change for 4.17 releaseSravan Kumar Lakkimsetti13-13/+13
Change-Id: Ic3961bfc4f546cbb07e403659abf419002af9250 Signed-off-by: Sravan Kumar Lakkimsetti <sravankumarl@in.ibm.com>
2020-05-28Bug 563372 - Be sure to handle async update properlyY20200609-2100Y20200609-0150S4_16_0_RC2R4_16I20200604-0540I20200603-0600I20200602-1800I20200602-0600I20200602-0510I20200602-0010I20200601-1800I20200601-0640I20200601-0140I20200531-1800I20200531-0600I20200530-0600I20200529-0550R4_16_maintenanceThomas Watson2-72/+43
testBug258209_1 has a timing issue because the Framework.update operation is async. This allows the test to try to waitForStop but the framework may have already be re-activated. Update this test to do what other tests that update the framework do by calling waitForStop in another thread before invoking Framework.update. Change-Id: Iee4de68f05b7cb151be2f61fcad353a625eb455c Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-21Bug 563372 - Capture thread dump for timeout during updateY20200528-0030S4_16_0_RC1I20200528-0200I20200528-0000I20200527-1800I20200527-0600I20200526-1800I20200526-0600I20200525-1800I20200525-0640I20200524-1800I20200524-0850I20200524-0600I20200524-0220I20200523-1800I20200523-0600I20200522-1800Thomas Watson2-3/+15
Change-Id: I96f1ef0e85ab873f56c612ff6e0d509bfd50a640 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-20Bug 563372 - Improve tests to capture why a framework does not stop.S4_16_0_M3I20200521-1000I20200520-1800Thomas Watson17-934/+284
Add a new method that captures a thread dump when the stop operation has a timeout. Change-Id: I938902904c1f972ac65a8d85db19b6e57a7ba580
2020-05-19Update OSGi R8 API to the latestThomas Watson9-151/+187
Change-Id: Ifb5c360e8c198a8def8065501136e1ef89598e4b Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-18Merge branch 'master' into osgiR8Thomas Watson4-11/+103
2020-05-18Bug 563304 - close event manager on STOPPED_SYSTEM_REFRESHEDI20200520-0600I20200519-1800I20200519-1130I20200519-1010I20200518-2220I20200518-1800Thomas Watson1-2/+3
The STOPPED_SYSTEM_REFRESHED event was new in the OSGi R7 specification but the code listening for framework STOPPED events to close the event manager was not updated to pay attention to the new event. The result is that when this new event is used during framework shutdown the event manager is not closed, causing a thread to remain active after framework stop. Change-Id: Ib1a1c360ede7ad1b733556bc2ce44c5469c984b2 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-15Bug 553773 - Avoid calling hooks while registeringI20200516-0600I20200515-1930Thomas Watson3-9/+100
This avoids getting the framework hook until after it has finished registering. Some hooks run into issues if they are allowed to be called while they are registering. For example, WeavingHooks that are registered by declarative services can cause issues with re-entering into activation by SCR. This change also avoids the constant get/unget service hooks registrations each time the hook is invoked. Hooks are high traffic services during event storms. This should improve performance when hooks are present. Change-Id: Ic937616d421f1d5944b614a91e74417070f06fd0 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-11Merge branch 'master' into osgiR8Thomas Watson23-59/+173
2020-05-11Update OSGi R8 API to the latestThomas Watson22-565/+112
Change-Id: Idb91df766682206851772a4d98e4abf9ab55631c Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-11Refactoring: Extract multiple usages of getter in loops to variableI20200513-1800I20200512-1800I20200511-1800Karsten Thoms1-20/+44
Extracted a variable where getters are accessed multiple times in a loop. determineRequiresVisibility: removed obsolete variable declaration. The variable raised a warning as it hides a field. Save action: Wildcard imports are replaced by regular imports. Also minor formatting changes. Change-Id: Ie5db8218368cf64bfb063ad7aad4bb3e1fc9b9a5 Signed-off-by: Karsten Thoms <karsten.thoms@karakun.com>
2020-05-08Bug 562980 - [osgi R8] Update ServiceReference.isAssignableTo withThomas Watson6-40/+84
clearification from R8 Change-Id: I7d9a78de8d0dc719ef33c5bde9401108acb2645a Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-08Bug 562971 - Make LogServiceManager an AllServiceListenerI20200510-1800I20200509-1800I20200509-0600I20200509-0450Thomas Watson1-2/+2
Change-Id: Iacdad5f6c5daa3c347cecaae22df304dcb501187 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-08Bug 560428 - Update versions for 4.16 release.Thomas Watson12-13/+13
Change-Id: I7f0b1b61e87ee4ff78ab2664b177090a5e44ab3f Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-08Bug 562907 - Assign zero length array constants to LazyData in settersKarsten Thoms3-19/+45
The given array arguments to setters are checked. If they are non-null and length 0, the existing zero length array constants are assigned to the fields of LazyData instead. Includes 4.16 version update for org.eclipse.osgi.compatibility.state Change-Id: Ia7b06db2c6b83ffb46a4a615971138726545575f Signed-off-by: Karsten Thoms <karsten.thoms@karakun.com>
2020-05-08Binaries v20200508-0906Eclipse Releng Bot6-6/+6
2020-05-05Bug 558514 - Add implied ServicePermission for get Condition serviceThomas Watson1-0/+2
Change-Id: If3575cabec696fb53ba13ddebad83bb0e6aec294 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-05-04Bug 562796: filter: Cache certain immutable valuesBJ Hargrave1-64/+23
We cache Long, Float, Double, Boolean, and Version objects created from the attribute value string in the filter expression to avoid repetitive parsing. Change-Id: I0f573db828b21a2472eddc831ac5e595a0c30d2b
2020-05-01Bug 562710 - filter out non-effective caps/reqs for fragmentsI20200506-1800I20200505-1800I20200504-1800I20200503-1800I20200502-1800I20200501-1800Thomas Watson2-1/+65
Change-Id: Ife170357e1714723cdcb913aa4ff4af188674cb4 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-04-28Bug 552573 - rename createBundleActivator to newBundleActivatorThomas Watson3-7/+7
Change-Id: I76646b16bc7ef2efab34d0747143b1a2e6c01197 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-04-13Disable org.eclipse.equinox.weaving.hooks.WeavingHook for testThomas Watson1-16/+19
The Equinox weaving hooks aggressively open all bundle files at framework start. Disable the hook to allow the testing of open/close properly. Change-Id: I3ed629c35c4135d44c08ece3ffa7151ed3707d3d
2020-04-02Merge branch 'master' into osgiR8Thomas Watson17-96/+303
2020-04-02Bug 558514 - [OSGi R8] Core changes for Condition Service RFC 242Thomas Watson6-8/+168
Change-Id: I7d27c1fe21d0bfcb4a621f5f5838b1cb029a82c3 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-04-01Bug 560788 - Improve runtime exceptions out of the frameworkS4_16_0_M1I20200421-1800I20200420-1800I20200419-1800I20200418-1800I20200417-1800I20200416-1800I20200416-0410I20200415-1800I20200415-0620I20200415-0200I20200414-1010I20200414-0630I20200413-1800I20200412-1800I20200411-1800I20200411-0950I20200411-0720I20200411-0510I20200411-0300I20200410-2300I20200410-1800I20200409-0200I20200408-1800I20200408-0600I20200407-1800I20200407-1210I20200407-0120I20200406-1800I20200406-0600I20200406-0350I20200406-0120I20200405-1800I20200405-0600I20200404-1800I20200404-1220I20200404-0600I20200404-0330I20200403-1800I20200402-2250I20200402-1800I20200402-0850I20200402-0410I20200401-1800Thomas Watson11-37/+203
Improve exception methods to include information about the bundle related to the exception. This is not an exhaustive change, but it does improve many of the common error that can happen when a bundle is uninstalled or a bundle context is invalid or a service is unregistered. Change-Id: I983a1fb4750f24259df71f73b4646134c3f1cc26 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-04-01Bug 561152 - Make move operation more robustThomas Watson6-59/+100
For staging files into the storage area use Files.move to try atomic moves. Fall back to non-atomic on failure. Also use replace option in case the target exists. Change-Id: Icd645470b25979dc76a7a1d9ea12b3e4837d634d Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-03-25Fix for null class context in ContextFinderThomas Watson1-1/+3
On Android SecurityManager.getClassContext returns null, this causes NPEs. Work around this by returning an empty array instead Change-Id: I0e3cda3b4fa11aae5bf840ffc68437d3e642e08c Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-03-25Update default profile to be Java 8Thomas Watson1-3/+10
Equinox is compiled to Java 8 so default to that profile if the current running Java version cannot be determined Change-Id: I92e834113ad08e5b779d428ed27ae02dd06e2581 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-03-12Merge branch 'master' into osgiR8Thomas Watson1-2/+9
2020-03-12Bug 561077 - set install area to user.dir if not loaded from file URLX20200313-0410X20200312-1900I20200323-1800I20200322-1800I20200322-1150I20200322-0300I20200321-1800I20200321-0440I20200320-1800I20200319-1800I20200319-0550I20200318-1800I20200318-1400I20200318-1100I20200317-1800I20200316-1800I20200315-2240I20200315-1800I20200314-1800I20200313-1800I20200313-0810I20200312-1835I20200312-1800Thomas Watson1-2/+9
Change-Id: I89aaf5894cf0b8546333f2b7dc1159733cd6cc2d Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-03-12Merge branch 'master' into osgiR8Thomas Watson1-1/+1
2020-03-12Bug 561041 - Use workaround to avoid errors on reflecting onX20200312-1530X20200312-1210Thomas Watson1-1/+1
URLClassLoader.addURL Change-Id: I960f40c5a3a268506d7046c0f9de7d4b54873a9b Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-03-10Update OSGi R8 APIsThomas Watson3-5/+17
2020-03-10Merge branch 'master' into osgiR8Thomas Watson15-69/+65
2020-03-10Bug 560974 - Fix exception type for null map keyBJ Hargrave2-56/+52
When a Map which does not support null keys receives a null key, the Map contract specified it should throw a NullPointerException. This class instead threw IllegalArgumentException. This change brings this class inline with the Map contract. Change-Id: Ia227aa18ac024f007dfc8ea1a2beaab12de393c2 Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
2020-03-10Bug 560974 - bump version of org.eclipse.osgi for 4.16Thomas Watson4-4/+4
Change-Id: I9c1360abb31123ab02666dca5471eba73a3a32a5 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2020-03-08Bug 560877 - Version bumps for 4.16 streamY20200309-1300I20200310-0810I20200309-1800I20200309-0810I20200308-1800Alexander Kurtakov2-2/+2
Change-Id: I176eb3632d5016378fb619bd99aef06fcadc9bf4 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2020-03-06Bug 560872 - POM and product version change for 4.16 releaseMat Booth13-13/+13
Change-Id: I9545bb9bdd28b5a1e5a0b7fcb6d645553fcec9d2 Signed-off-by: Mat Booth <mat.booth@redhat.com>
2020-02-21Bug 559840 - handle duplicate ConnectContent with ConnectBundleFileThomas Watson3-2/+33
We need to make sure we only create one ConnectBundleFile instance per ConnectContent instance. Otherwise our managing of open/close operations will get messed up. Change-Id: I6fa3f9857726a39eaee8e07f7e11abdee9881abb
2020-02-21Bug 559840 - handle connect content for system bundle correctlyThomas Watson3-43/+100
Change-Id: I1809c5fe464b23b445eb4fbd42f5c318dff48c6b
2020-02-19Merge branch 'master' into osgiR8Thomas Watson418-2853/+3003
2020-02-15Convert osgi.tests to JUnit 4.Y20200409-0600Y20200407-0925Y20200407-0635Y20200407-0550Y20200318-0540Y20200317-0225Y20200316-1300Y20200316-0820Y20200314-1405Y20200313-1100Y20200311-0915Y20200310-0525Y20200305-0600Y20200225-1040Y20200219-1055Y20200219-0850Y20200218-1420Y20200218-1210Y20200218-0315Y20200218-0120X20200306-0700X20200305-1900X20200305-0700X20200304-1900X20200304-0700X20200303-1900X20200303-0700X20200302-1900X20200302-0930X20200302-0700X20200301-1900X20200301-0700X20200229-1900X20200229-0700X20200228-1900X20200228-0700X20200228-0520X20200227-1900X20200227-0700X20200227-0130X20200226-0700X20200225-1900X20200225-0700X20200224-1900X20200224-0700X20200223-1900X20200223-0700X20200223-0250X20200222-0700X20200221-1130X20200221-0700X20200221-0440X20200221-0310X20200221-0050X20200220-1900X20200220-0700X20200219-1900X20200219-0700X20200219-0500X20200218-0510X20200217-1120S4_15_0_RC2S4_15_0_RC1S4_15_0_M3R4_15I20200305-0155I20200304-1800I20200304-0600I20200303-1800I20200303-0600I20200302-1800I20200302-0600I20200301-2310I20200301-1800I20200301-0600I20200301-0030I20200229-1800I20200229-0600I20200228-1800I20200226-1800I20200226-0600I20200225-1800I20200225-0600I20200224-1800I20200224-0600I20200223-2335I20200223-1800I20200223-0600I20200222-1800I20200222-0600I20200221-1800I20200219-1800I20200219-0600I20200218-1800I20200218-0600I20200217-1800I20200217-0600I20200216-1800I20200216-1145I20200215-1800Alexander Kurtakov34-523/+613
Had to move to default eclipse formatter as the ancient one used simply had no idea what to do with annotations. Change-Id: I3dc61b31a2a5edf0da03ee17ed03d241c1178607 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2020-02-14Bug 559829 - [OSGi R8] connect tags for osgi.connect must propagate toAnjum Fatima2-9/+23
other BSN capabilities Change-Id: I9e18599cf4add93626e98d4cce6241cf9258619f Signed-off-by: Anjum Fatima <anjum.eclipse@gmail.com>
2020-02-14Bug 559118 - Don't fail with runtime exception when substitutes are notI20200215-0600I20200214-1800Thomas Watson1-5/+13
handled correctly This doesn't impact the framework, but the resolver is also registered as a service and used by others. The original fix would cause a runtime exception if no sources could be found for a blamed used capability. The previous resolver implementation would silently ignore the source and it would not impact the used class space. This still produces a valid class space according to what the resolve context is giving the resolver to work with. We will just log an info message if this happens. Change-Id: I34c61efa33a703c66763197146d09a62f09f9ad8
2020-02-12Bug 560070 - [resolver] resolving in batch can override substitedI20200214-0620I20200213-1800I20200213-1520I20200213-1210I20200213-1100I20200213-0640Thomas Watson4-14/+86
packages Change-Id: Ieb07bfa4d7ba17cc22859f1ed4f12889e23efbfe Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>

Back to the top