Skip to main content
aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2017-05-02Bug 516069 - reflection code to create permission objects should beI20170502-2000Thomas Watson1-6/+18
protected with doPriv Change-Id: Id08f3a33f835b2eb1e9ad355126a540307588be4 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-04-27Bug 515920 - "Windows Server 2016" to osname.aliases I20170428-0550I20170428-0445I20170427-2000Thomas Watson2-1/+8
Change-Id: I07d5537eddb796db7a434a3616bc50acdc51b8ec Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-04-14Bug 515286 - Init of FrameworkExtensionInstaller fails with JDK9Thomas Watson1-2/+7
Change-Id: Id0544936abf98b465ea8cd8bf1558d7c19d6e379 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-04-05Bug 514802 - Return empty fragment constant when fragmentWiresKarsten Thoms1-1/+1
collection is empty Signed-off-by: Karsten Thoms <karsten.thoms@itemis.de>
2017-04-04Bug 508357 - Native library is not loaded from bundle on SolarisI20170405-0300I20170405-0200I20170405-0130I20170404-2000Thomas Watson1-2/+2
Make SunOS and Solaris alias each other. Change-Id: Ic40e4d7888530fa328eaea2460db45225a393a67 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-03-30Revert "Bug 502220 - Default workspace is ${launcher.name}-workspace"I20170402-2000I20170401-2000I20170331-2000I20170330-2000Thomas Watson1-15/+2
This reverts commit 60a423633ebd7246d33e8f95eb642e1dbeb2ea07. Change-Id: I7aad1e0ff3d3e870595fe4fa3883451dce434b89
2017-03-29Bug 502220 - Default workspace is ${launcher.name}-workspaceMickael Istria1-2/+15
Change-Id: If2d885a7145222e626813e3b29a395f9adef46b3 Signed-off-by: Mickael Istria <mistria@redhat.com>
2017-03-02Bug 512707 - Add support for recursive class processingY20170321-2315Y20170321-0120Y20170316-0500Y20170309-1000S4_7_0_M6I20170327-2000I20170326-2000I20170325-2000I20170324-2000I20170323-2000I20170322-2000I20170321-2000I20170320-2000I20170319-2000I20170318-2000I20170317-2000I20170316-2000I20170315-2000I20170315-0500I20170314-2000I20170313-2000I20170312-2000I20170311-2000I20170310-2000I20170308-2000I20170308-0800I20170307-2000I20170307-0800I20170307-0610I20170307-0450I20170307-0420I20170307-0410I20170307-0150I20170306-2000I20170306-0800I20170305-2000I20170304-2000I20170303-2000I20170302-2000Thomas Watson12-63/+230
Change-Id: I3223216586d1cb5bf399132b26293cf0e787377f Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-02-09Bug 511593 - Issue resolving split packagesY20170223-1000Y20170216-1515Y20170216-1450Y20170216-1000I20170228-0330I20170228-0240I20170227-2000I20170226-2000I20170225-2000I20170224-2000I20170223-2000I20170222-2000I20170221-2000I20170220-2000I20170219-2000I20170218-2000I20170217-2000I20170216-2000I20170216-0140I20170215-2000I20170214-2200I20170214-2000I20170213-2000I20170212-2000I20170211-2000I20170210-2000I20170210-0615I20170210-0600I20170210-0550I20170210-0530I20170209-2000Thomas Watson2-18/+7
Change solution to avoid introducing new API on ModuleWiring. Change-Id: I1de1dd5ce55c05b04f9eaa699606ce70ba22ec89 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-02-09Bug 511593 - Issue resolving split packagesThomas Watson1-0/+11
Add javadoc for the getSubstitutionWires method Change-Id: I10f81b743365eda2a38b8a08e4e8547612a90e70 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-02-03Bug 511268 - Tracing should honour -consoleLogY20170209-1000I20170208-2000I20170208-0700I20170207-2000I20170206-2000Thomas Watson1-13/+26
Change-Id: Ib7ce9ab914f96f1ec739c8e88be8aaf4b6c269f2 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-02-03Bug 511593 - Issue resolving split packagesI20170205-2000I20170204-2000I20170203-2000Thomas Watson1-2/+25
The resolver implementation fix. Change-Id: I71384aec4decc0a7b009c5d48815ade3a8a4c010 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-01-18Restore Headers temporarily.Y20170202-1000Y20170126-1700Y20170126-1000Y20170119-1000S4_7_0_M5I20170202-2000I20170201-2000I20170131-2000I20170130-2000I20170130-0600I20170129-2000I20170128-2000I20170127-2200I20170127-2000I20170126-1030I20170126-0505I20170125-2000I20170125-0800I20170124-2000I20170124-0800I20170123-2000I20170123-0830I20170123-0800I20170122-2000I20170121-2000I20170121-0950I20170120-2000I20170119-2000I20170119-1010I20170119-0345I20170118-2000Thomas Watson1-0/+338
Change-Id: I202447a6279febe848c5db0b2c301650b06274f6 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-01-18Bug 510641 - Remove old Headers implementation.Thomas Watson1-338/+0
Change-Id: Ibc09bc478f68516c692c95181081a08781024ed5 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2017-01-18Bug 510641 - Remove CaseInsensitiveDictionaryMap.putIfAbsent methodBJ Hargrave2-64/+18
We don’t actually need it since put can be used in the constructor when checking for case-variant key issues. Removing putIfAbsent then simplifies some implementation. Change-Id: Id8ae7a2b90b0edd3662f8697d64734e314e20cf8 Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
2017-01-18Bug 510641 - Alternate CaseInsentiveDictionaryMap implementationBJ Hargrave1-34/+255
This implementation should provide for O(1) lookup even for String key misses. This is done by wrapping String keys in a case-insensitive wrapper with each string key wrapped by a new wrapper object. Change-Id: I083a5141764c3c1108ff72afcb76918e27296bd4 Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
2017-01-18Bug 510641 - service registry: Rewrite ServicePropertiesBJ Hargrave4-119/+98
The class is now based on CaseInsensitiveDictionaryMap (instead of Headers). Once created, the service properties are internally represented as a Map. Change-Id: I3041237d73e36a533c6d73985d75fb680ad129d7 Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
2017-01-18Bug 510641 - service registry: Use StringBuilder instead of StringBufferBJ Hargrave2-4/+4
Change-Id: I222c6a298d1ad5fb3aa19c3f122f5c4d8dd97636 Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
2017-01-18Bug 510641 - Covert use of Headers to CaseInsensitiveDictionaryMapBJ Hargrave4-29/+29
Change-Id: If786a84bf13039cf92204dad8b0afe4abc75c9af Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
2017-01-18Bug 510641 - headers: Replace Headers with CaseInsensitiveDictionaryMapBJ Hargrave2-1/+413
The CaseInsensitiveDictionaryMap class uses an internal map object. This allows for O(1) lookup for non-String keys and for String keys if the original key case is used; and for O(log n) if a String case-variant is used. The map implementation is complete. Change-Id: I43a4b96d2dca4b4057762f22f98d17db923d67aa Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
2016-11-17Bug 507703 - Add toString to ModuleWiringY20161124-1000I20161128-2000I20161128-0600I20161128-0330I20161127-2000I20161127-1300I20161126-2000I20161126-0815I20161125-2000I20161124-2000I20161123-2000I20161122-2000I20161121-2000I20161120-2000I20161119-2000I20161118-2000I20161117-2000Thomas Watson1-0/+5
Change-Id: I4067437e5da201f472c48173a2595e63bb589177 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-11-11Bug 507417 [java 9] cannot load java.* classes that are not contained inY20161117-1000I20161116-2000I20161115-2000I20161114-2000I20161114-0355I20161113-2000I20161112-2000I20161112-0405I20161111-2000Thomas Watson4-13/+20
the java.base module Change-Id: I788208133cd71e4ce6bbdc92bea890a8cf7d5ce4 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-11-07Bug 507178 - Issues with encoding/decoding file URLs in variousY20161110-1000I20161111-1220I20161111-0900I20161111-0740I20161111-0510I20161110-2030I20161110-1630I20161110-1225I20161110-1130I20161110-0710I20161110-0510I20161109-2000I20161108-1015I20161108-0800Thomas Watson9-101/+49
scenarios Change-Id: I2cf6b79d055eef7fb180f3f0f75032cbb84b7d51 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-11-07Bug 502209 - remove printStackTraceThomas Watson1-1/+0
Change-Id: Iacfe8995f5a3479c9c4cb947efcfb6fb7dabd6df Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-10-27Bug 502209 - Reflection used for URL handler support breaks on Java 9Thomas Watson5-17/+150
Change-Id: I8069523103f1e3966feb9856e2a1100eda6b02b2 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-10-27Bug 506578 - Custom class loaders mapped to Java 9 modules must overrideThomas Watson1-0/+14
new findResource and findClass methods Change-Id: I25ab9c6a154314216e9c6ff492b4062f3f4bba42 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-10-25Bug 506460 - openConnection(URL u, Proxy p) is unsupported when multipleI20161025-2000Thomas Watson2-1/+25
frameworks are running Change-Id: I44b31638002212d08a0a47bd4d9511d88420cc3b Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-10-20Bug 506208 - Race to call hooks inY20161024-2045Y20161024-1300Y20161024-0700Y20161021-1200Y20161021-0600Y20161020-1800I20161024-2000I20161024-1000I20161024-0410I20161023-2000Thomas Watson1-14/+28
org.eclipse.osgi.internal.loader.BundleLoader.getModuleClassLoader() Change-Id: I6ec623901997754052119b1a2cf3880678bda05a Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-10-04Bug 503742 - Disable by default lazy start trigger on failed class loadY20161018-1100Y20161013-1000Y20161006-1000I20161018-0800I20161011-0800I20161005-1430I20161005-1045I20161005-0815Thomas Watson2-2/+6
from Bundle.loadClass Change-Id: I4d9533461b4ed4db76dc96772fd4a50a9d28a1c0 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-10-03Bug 502204 - Reflection used for parallel capable class loaders willThomas Watson2-28/+3
fail on Java 9 Change-Id: I10ee55af8f1844f314f0a466eaf51c4ce7b99da8 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-09-29Bug 502425 - Refactor code for Java 7Thomas Watson92-581/+570
Change-Id: I64d3a28a743a0292db27fd421294421dc51eb2e1 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-09-29Bug 502425 - Move Equinox Framework up to Java 7Thomas Watson1-3/+3
Change-Id: I2dd55f8926f6d3fe4c15c664b0cc839545711bc1 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-09-27Bug 502318 - osgi.framework.activeThreadType is no longer usedY20160929-1000Thomas Watson2-2/+10
Change-Id: I54fa1cce33e778d02a792b7c24a651c30555d08e Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-09-26Bug 502108 - Initialization of FrameworkExtensionInstaller fails withI20160927-0800Thomas Watson1-1/+5
JDK9 Change-Id: Idf30e3c56eff1b8ae48b7138539e82da150c4af6 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-09-08Bug 501063 - Use constructor in favor of Throwable.initCause(...)Y20160908-1000I20160913-0800I20160912-2000I20160912-1005I20160912-0800I20160912-0430I20160912-0225I20160912-0140I20160911-2015Andreas Sewe9-16/+14
Change-Id: I8e8dde9b777054ea62c5e6d0ce38191d50bf8c92 Signed-off-by: Andreas Sewe <andreas.sewe@codetrails.com>
2016-08-19Bug 499971 - Enrich IOExceptions thrown by ZipBundleEntryInputStreamY20160825-1000I20160830-0800I20160824-1429I20160823-1359I20160823-0759Andreas Sewe1-18/+40
To aid debugging, this wraps any IOExceptions (e.g., ZIPExceptions) thrown by the underlying (Zip)InputStream in IOExceptions indicating which file is broken. Change-Id: I21d447b3aafa36b3d83a1f22bcc2fe776d6f9737 Signed-off-by: Andreas Sewe <andreas.sewe@codetrails.com>
2016-08-12Bug 499617 - Avoid holding class loader locks while lazy activatingY20160818-1000I20160816-1015I20160816-0800Thomas Watson1-30/+26
Change-Id: Ided8055478bcc15c6669bd665b29a0c88e937728 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-08-12Bug 497576 - [equinox framework] augmenting bundle metadataThomas Watson6-16/+87
Container changes to allow a container adaptor to adapt a builder before it is used by the container. EquinoxContainerAdaptor changes to call StorageHook implementations to adapt a builder for a generation. Change-Id: I4597f98b6fc6c993fc601a615f61047a8299beae Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-07-13Bug 486952 - [osgi R7] set org.osgi.framework.storage property whenY20160811-1000Y20160804-1000Y20160728-1000Y20160721-1000Y20160714-1000I20160809-1300I20160809-1100I20160803-2000I20160803-0800I20160802-2000I20160802-0800I20160801-2000I20160801-0800I20160731-2000I20160726-1400I20160726-0800I20160719-0800I20160716-1900Thomas Watson1-0/+6
derived Change-Id: I7ff18cea64a58827ab74374c2b8db5fb97b80837 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-07-08Bug 416164 - Correctly handle spurious wake ups.Thomas Watson1-42/+59
Replace forStop may with a single AtomicReference that we can hand out to waiting threads and notify when the framework stops Change-Id: I1f6da6b9c896ba1bc2397bccfd9b34899a497bce Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-07-06Bug 497407 - Handle @initial locations for bundle update operationsY20160707-1000Thomas Watson1-1/+8
Change-Id: I396cf341934c0de5f0b589ba691a913f0745766a Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-06-30Bug 497092 - Use INFO event from uninstalled/updated bundle revisionI20160705-1600Thomas Watson1-1/+9
If trying to open a zip file and an IOException is thrown then check if the bundle revision still exists in the list of revisions for the bundle. If it does not then the revision was removed as part of an uninstall or update operation. In this case just use an INFO event. Change-Id: I975061a2e963be1c7b9752401aae6d70eddda003 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-06-24Bug 496034 - hold no locks while creating ModuleClassLoadersThomas Watson1-43/+57
Change-Id: I0a16d4fd7b7ff5dbdb862b555612af5a7d3d7482 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-06-24Revert "Bug 496034 - Obtain readLock on ModuleDatabase before ↵Thomas Watson1-37/+13
classLoaderMonitor" This reverts commit 171aebbb2152cf89061c8618dd836341dbd83530. Change-Id: Ic4fb1b6c96ca01f61cc357a58e0f37cfd25ab585
2016-06-16Bug 496259 - Remove implementation of UniversalUniqueIdentifierThomas Watson3-272/+2
Change-Id: I1d72a774f7cde0dd9942aa2af00b506352f3a92f Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-06-16Revert "Bug 449765 - Framework should provide a persistent UUID"Thomas Watson5-64/+273
This reverts commit 3841a41167128ab80ffe8f05bf793694f5c28400. Change-Id: I90c3d843d6593997f1ff38922a3662d2f4131509
2016-06-15Bug 449765 - Framework should provide a persistent UUIDThomas Watson5-273/+64
Change-Id: Ic26c63aa5fbec649786070447c737bd81cf5521d Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-06-14Bug 496034 - Obtain readLock on ModuleDatabase before classLoaderMonitorThomas Watson1-13/+37
To prevent potential out of order locks between the ModuleDatabase read/write lock and the classLoaderMonitor lock we obtain the readLock on the ModuleDatabase before the classLoaderMonitor. This is because during the refresh operation the removed wirings will close the BundleLoader while holding the writeLock, this eventually nests a lock with the classLoaderMonitor monitor lock. Since we call foreign hook code while holding the classLoaderMonitor monitor lock we should be consistent in the order locks are obtained to prevent potential deadlocks by the hooks. Change-Id: I0987756b2046ca8d8076bc7ad01a2cc9e87e18ff Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-06-13Bug 493387 - Resolver timeout can be improved by overriding the LoggerThomas Watson1-2/+18
used by the resolver Change-Id: I2722ac5512e070a3fbb3e4da0e6c527c7eb1552c Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>
2016-06-03Bug 495333 - ContextFinder returns null from getResourcesR4_6I20160606-1100I20160603-1000Thomas Watson1-3/+12
If there are no classes loaded from bundle class loaders on the stack then ContextFinder.getResources can return null. This fix ensures an empty enumeration is returned instead. Change-Id: I07fe25eeaf8d02ea92cbe130794cdd2a011ae481 Signed-off-by: Thomas Watson <tjwatson@us.ibm.com>

Back to the top