158170 2016-05-27 15:20:44 -0700 DFG::LazyJSValue::tryGetStringImpl() crashes for empty values 2016-05-28 06:00:19 -0700 1 1 1 Unclassified WebKit JavaScriptCore WebKit Nightly Build All All RESOLVED FIXED P2 Normal --- 1 fpizlo fpizlo commit-queue keith_miller mark.lam msaboff saam oldest_to_newest 1197236 0 fpizlo 2016-05-27 15:20:44 -0700 Patch forthcoming. 1197238 1 280002 fpizlo 2016-05-27 15:23:09 -0700 Created attachment 280002 the patch 1197241 2 280002 msaboff 2016-05-27 15:25:45 -0700 Comment on attachment 280002 the patch r=me 1197244 3 fpizlo 2016-05-27 15:28:16 -0700 Landed in http://trac.webkit.org/changeset/201470 1197334 4 280002 saam 2016-05-27 23:33:17 -0700 Comment on attachment 280002 the patch View in context: https://bugs.webkit.org/attachment.cgi?id=280002&action=review > Source/JavaScriptCore/ChangeLog:9 > + value, presumably because this makes it soooper fast. In DFG IR, empty values can appear I'm going to run the experiment where jsDynamicCast<Foo*>(JSValue()) returns nullptr and see what the performance is like. Do you know if anyone has measured the performance impact of this recently? 1197369 5 fpizlo 2016-05-28 06:00:19 -0700 (In reply to comment #4) > Comment on attachment 280002 [details] > the patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=280002&action=review > > > Source/JavaScriptCore/ChangeLog:9 > > + value, presumably because this makes it soooper fast. In DFG IR, empty values can appear > > I'm going to run the experiment where jsDynamicCast<Foo*>(JSValue()) returns > nullptr and see what the performance is like. > Do you know if anyone has measured the performance impact of this recently? I don't think anyone has ever measured it! 280002 2016-05-27 15:23:09 -0700 2016-05-27 15:25:45 -0700 the patch blah.patch text/plain 2930 fpizlo SW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9DaGFuZ2VMb2cKPT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291 cmNlL0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkocmV2aXNpb24gMjAxNDY5KQorKysgU291cmNl L0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDI2IEBA CisyMDE2LTA1LTI3ICBGaWxpcCBQaXpsbyAgPGZwaXpsb0BhcHBsZS5jb20+CisKKyAgICAgICAg REZHOjpMYXp5SlNWYWx1ZTo6dHJ5R2V0U3RyaW5nSW1wbCgpIGNyYXNoZXMgZm9yIGVtcHR5IHZh bHVlcworICAgICAgICBodHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93X2J1Zy5jZ2k/aWQ9MTU4 MTcwCisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgVGhl IHByb2JsZW0gaGVyZSBpcyB0aGF0IGpzRHluYW1pY0Nhc3Q8PigpIGlzIGV2aWwhIEl0IGF2b2lk cyBjaGVja2luZyBmb3IgdGhlIGVtcHR5CisgICAgICAgIHZhbHVlLCBwcmVzdW1hYmx5IGJlY2F1 c2UgdGhpcyBtYWtlcyBpdCBzb29vcGVyIGZhc3QuIEluIERGRyBJUiwgZW1wdHkgdmFsdWVzIGNh biBhcHBlYXIKKyAgICAgICAgYW55d2hlcmUgYmVjYXVzZSBvZiBURFouCisgICAgICAgIAorICAg ICAgICBUaGlzIHBhdGNoIGRvZXNuJ3QgY2hhbmdlIGpzRHluYW1pY0Nhc3Q8PigpLCBidXQgaXQg aGFyZGVucyBvdXIgd3JhcHBlcnMgZm9yIGl0IGluIHRoZSBERkcKKyAgICAgICAgYW5kIGl0IGhh cyB0aGUgYWZmZWN0ZWQgY29kZSB1c2Ugb25lIG9mIHRob3NlIHdyYXBwZXJzLgorICAgICAgICAK KyAgICAgICAgKiBkZmcvREZHRnJvemVuVmFsdWUuaDoKKyAgICAgICAgKEpTQzo6REZHOjpGcm96 ZW5WYWx1ZTo6ZHluYW1pY0Nhc3QpOiBIYXJkZW4gdGhpcy4KKyAgICAgICAgKEpTQzo6REZHOjpG cm96ZW5WYWx1ZTo6Y2FzdCk6CisgICAgICAgICogZGZnL0RGR0xhenlKU1ZhbHVlLmNwcDoKKyAg ICAgICAgKEpTQzo6REZHOjpMYXp5SlNWYWx1ZTo6dHJ5R2V0U3RyaW5nSW1wbCk6IFVzZSB0aGUg aGFyZGVuZWQgd3JhcHBlci4KKyAgICAgICAgKiB0ZXN0cy9zdHJlc3Mvc3RyY2F0LWVtdHB5Lmpz OiBBZGRlZC4gVGhpcyB1c2VkIHRvIGNyYXNoIGV2ZXJ5IHRpbWUuCisgICAgICAgIChmb28pOgor ICAgICAgICAoaS5jYXRjaCk6CisKIDIwMTYtMDUtMjcgIEZpbGlwIFBpemxvICA8ZnBpemxvQGFw cGxlLmNvbT4KIAogICAgICAgICByZWdFeHBQcm90b0Z1bmNTcGxpdEZhc3Qgc2hvdWxkIE9PTSBi ZWZvcmUgaXQgc3dhcHMKSW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9kZmcvREZHRnJvemVu VmFsdWUuaAo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvSmF2YVNjcmlwdENvcmUvZGZnL0RGR0Zyb3pl blZhbHVlLmgJKHJldmlzaW9uIDIwMTQ2OSkKKysrIFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9kZmcv REZHRnJvemVuVmFsdWUuaAkod29ya2luZyBjb3B5KQpAQCAtNzMsNyArNzMsMTAgQEAgcHVibGlj OgogICAgIHRlbXBsYXRlPHR5cGVuYW1lIFQ+CiAgICAgVCBkeW5hbWljQ2FzdCgpCiAgICAgewot ICAgICAgICByZXR1cm4ganNEeW5hbWljQ2FzdDxUPih2YWx1ZSgpKTsKKyAgICAgICAgSlNWYWx1 ZSB0aGVWYWx1ZSA9IHZhbHVlKCk7CisgICAgICAgIGlmICghdGhlVmFsdWUpCisgICAgICAgICAg ICByZXR1cm4gbnVsbHB0cjsKKyAgICAgICAgcmV0dXJuIGpzRHluYW1pY0Nhc3Q8VD4odGhlVmFs dWUpOwogICAgIH0KICAgICB0ZW1wbGF0ZTx0eXBlbmFtZSBUPgogICAgIFQgY2FzdCgpCkluZGV4 OiBTb3VyY2UvSmF2YVNjcmlwdENvcmUvZGZnL0RGR0xhenlKU1ZhbHVlLmNwcAo9PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 Ci0tLSBTb3VyY2UvSmF2YVNjcmlwdENvcmUvZGZnL0RGR0xhenlKU1ZhbHVlLmNwcAkocmV2aXNp b24gMjAxNDY5KQorKysgU291cmNlL0phdmFTY3JpcHRDb3JlL2RmZy9ERkdMYXp5SlNWYWx1ZS5j cHAJKHdvcmtpbmcgY29weSkKQEAgLTk1LDcgKzk1LDcgQEAgY29uc3QgU3RyaW5nSW1wbCogTGF6 eUpTVmFsdWU6OnRyeUdldFN0cgogICAgICAgICByZXR1cm4gdS5zdHJpbmdJbXBsOwogCiAgICAg Y2FzZSBLbm93blZhbHVlOgotICAgICAgICBpZiAoSlNTdHJpbmcqIHN0cmluZyA9IGpzRHluYW1p Y0Nhc3Q8SlNTdHJpbmcqPih2YWx1ZSgpLT52YWx1ZSgpKSkKKyAgICAgICAgaWYgKEpTU3RyaW5n KiBzdHJpbmcgPSB2YWx1ZSgpLT5keW5hbWljQ2FzdDxKU1N0cmluZyo+KCkpCiAgICAgICAgICAg ICByZXR1cm4gc3RyaW5nLT50cnlHZXRWYWx1ZUltcGwoKTsKICAgICAgICAgcmV0dXJuIG51bGxw dHI7CiAKSW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS90ZXN0cy9zdHJlc3Mvc3RyY2F0LWVt dHB5LmpzCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT0KLS0tIFNvdXJjZS9KYXZhU2NyaXB0Q29yZS90ZXN0cy9zdHJlc3Mv c3RyY2F0LWVtdHB5LmpzCShyZXZpc2lvbiAwKQorKysgU291cmNlL0phdmFTY3JpcHRDb3JlL3Rl c3RzL3N0cmVzcy9zdHJjYXQtZW10cHkuanMJKHdvcmtpbmcgY29weSkKQEAgLTAsMCArMSwxNCBA QAorZnVuY3Rpb24gZm9vKCkgeworICAgICJ1c2Ugc3RyaWN0IjsKKyAgICBsZXQgYSA9ICJoZWxs byIgKyBhOworICAgIHJldHVybiBhOworfQorCitub0lubGluZShmb28pOworCitmb3IgKHZhciBp ID0gMDsgaSA8IDEwMDAwOyArK2kpIHsKKyAgICB0cnkgeworICAgICAgICBmb28oKTsKKyAgICB9 IGNhdGNoIChlKSB7CisgICAgfQorfQo=