Also fix Heap's extra memory size accounting to handle overflows. <rdar://problem/29697336>
Created attachment 306104 [details] proposed patch.
Comment on attachment 306104 [details] proposed patch. Attachment 306104 [details] did not pass mac-debug-ews (mac): Output: http://webkit-queues.webkit.org/results/3467174 Number of test failures exceeded the failure limit.
Created attachment 306115 [details] Archive of layout-test-results from ews115 for mac-elcapitan The attached test failures were seen while running run-webkit-tests on the mac-debug-ews. Bot: ews115 Port: mac-elcapitan Platform: Mac OS X 10.11.6
Comment on attachment 306104 [details] proposed patch. Attachment 306104 [details] did not pass jsc-ews (mac): Output: http://webkit-queues.webkit.org/results/3467269 New failing tests: stress/recursive-try-catch.js.ftl-no-cjit-small-pool stress/recursive-try-catch.js.dfg-eager-no-cjit-validate stress/recursive-try-catch.js.no-ftl stress/recursive-try-catch.js.dfg-maximal-flush-validate-no-cjit stress/recursive-try-catch.js.no-cjit-validate-phases stress/recursive-try-catch.js.ftl-no-cjit-no-inline-validate stress/recursive-try-catch.js.dfg-eager stress/recursive-try-catch.js.ftl-no-cjit-no-put-stack-validate stress/recursive-try-catch.js.no-cjit-collect-continuously stress/recursive-try-catch.js.ftl-eager-no-cjit stress/recursive-try-catch.js.default stress/recursive-try-catch.js.no-llint stress/new-largeish-contiguous-array-with-size.js.ram-size-10000000 stress/recursive-try-catch.js.ftl-eager
Created attachment 306120 [details] Patch for landing: w/ 1 fix. In Heap::extraMemorySize(), I meant to take the min of (total, std::numeric_limits<size_t>::max() - m_objectSpace.capacity()), not the max. This is the reason for the JSC test failure and possibly the layout test failures too. Let's try this on the EWS bots.
Comment on attachment 306120 [details] Patch for landing: w/ 1 fix. This is the kind of change where the test is usually more valuable than the code change. Can you think of a way to test this?
(In reply to Geoffrey Garen from comment #6) > Comment on attachment 306120 [details] > Patch for landing: w/ 1 fix. > > This is the kind of change where the test is usually more valuable than the > code change. Can you think of a way to test this? I found a way to make a test. Will add it to the patch.
Created attachment 306145 [details] patch for landing w/ test case.
Thanks for the review. Landed in r214857: <http://trac.webkit.org/r214857>.