Title: [249417] trunk/Source/WebCore
- Revision
- 249417
- Author
- ysuz...@apple.com
- Date
- 2019-09-02 22:59:35 -0700 (Mon, 02 Sep 2019)
Log Message
[WebCore] Resource usage accounting should accept non KERN_SUCCESS
https://bugs.webkit.org/show_bug.cgi?id=201409
Reviewed by Andreas Kling.
While iterating threads, we are not suspending these threads. Underlying threads can have gone
at any time and we will get non KERN_SUCCESS error code when a thread has gone. We should ignore
these threads.
* page/cocoa/ResourceUsageThreadCocoa.mm:
(WebCore::threadInfos):
Modified Paths
Diff
Modified: trunk/Source/WebCore/ChangeLog (249416 => 249417)
--- trunk/Source/WebCore/ChangeLog 2019-09-03 01:55:27 UTC (rev 249416)
+++ trunk/Source/WebCore/ChangeLog 2019-09-03 05:59:35 UTC (rev 249417)
@@ -1,3 +1,17 @@
+2019-09-02 Yusuke Suzuki <ysuz...@apple.com>
+
+ [WebCore] Resource usage accounting should accept non KERN_SUCCESS
+ https://bugs.webkit.org/show_bug.cgi?id=201409
+
+ Reviewed by Andreas Kling.
+
+ While iterating threads, we are not suspending these threads. Underlying threads can have gone
+ at any time and we will get non KERN_SUCCESS error code when a thread has gone. We should ignore
+ these threads.
+
+ * page/cocoa/ResourceUsageThreadCocoa.mm:
+ (WebCore::threadInfos):
+
2019-09-02 Fujii Hironori <hironori.fu...@sony.com>
[SVG] fragment-only url 'url(#fragment)' should be resolved against the current document with regardless to HTML <base> element
Modified: trunk/Source/WebCore/page/cocoa/ResourceUsageThreadCocoa.mm (249416 => 249417)
--- trunk/Source/WebCore/page/cocoa/ResourceUsageThreadCocoa.mm 2019-09-03 01:55:27 UTC (rev 249416)
+++ trunk/Source/WebCore/page/cocoa/ResourceUsageThreadCocoa.mm 2019-09-03 05:59:35 UTC (rev 249417)
@@ -87,6 +87,8 @@
if (kr != KERN_SUCCESS)
return { };
+ // Since we collect thread usage without suspending threads, threads retrieved by task_threads can have gone while iterating
+ // them to get usage information. If a thread has gone, a system call returns non KERN_SUCCESS, and we just ignore these threads.
Vector<ThreadInfo> infos;
for (mach_msg_type_number_t i = 0; i < threadCount; ++i) {
MachSendRight sendRight = MachSendRight::adopt(threadList[i]);
@@ -94,7 +96,6 @@
thread_info_data_t threadInfo;
mach_msg_type_number_t threadInfoCount = THREAD_INFO_MAX;
kr = thread_info(sendRight.sendRight(), THREAD_BASIC_INFO, reinterpret_cast<thread_info_t>(&threadInfo), &threadInfoCount);
- ASSERT(kr == KERN_SUCCESS);
if (kr != KERN_SUCCESS)
continue;
@@ -101,7 +102,6 @@
thread_identifier_info_data_t threadIdentifierInfo;
mach_msg_type_number_t threadIdentifierInfoCount = THREAD_IDENTIFIER_INFO_COUNT;
kr = thread_info(sendRight.sendRight(), THREAD_IDENTIFIER_INFO, reinterpret_cast<thread_info_t>(&threadIdentifierInfo), &threadIdentifierInfoCount);
- ASSERT(kr == KERN_SUCCESS);
if (kr != KERN_SUCCESS)
continue;
@@ -108,7 +108,6 @@
thread_extended_info_data_t threadExtendedInfo;
mach_msg_type_number_t threadExtendedInfoCount = THREAD_EXTENDED_INFO_COUNT;
kr = thread_info(sendRight.sendRight(), THREAD_EXTENDED_INFO, reinterpret_cast<thread_info_t>(&threadExtendedInfo), &threadExtendedInfoCount);
- ASSERT(kr == KERN_SUCCESS);
if (kr != KERN_SUCCESS)
continue;
@@ -117,6 +116,7 @@
if (!(threadBasicInfo->flags & TH_FLAGS_IDLE))
usage = threadBasicInfo->cpu_usage / static_cast<float>(TH_USAGE_SCALE) * 100.0;
+ // FIXME: dispatch_queue_t can be destroyed concurrently while we are accessing to it here. We should not use it.
String threadName = String(threadExtendedInfo.pth_name);
String dispatchQueueName;
if (threadIdentifierInfo.dispatch_qaddr) {
_______________________________________________
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
