131449 2014-04-09 12:34:37 -0700 Ensure that LLINT accessing of the ProtoCallFrame is big endian friendly 2014-04-09 14:53:45 -0700 1 1 1 Unclassified WebKit JavaScriptCore 528+ (Nightly build) Unspecified Unspecified RESOLVED FIXED P2 Normal --- 128743 1 mark.lam mark.lam fpizlo ggaren mhahnenberg mmirman msaboff oliver tpopela oldest_to_newest 999256 0 mark.lam 2014-04-09 12:34:37 -0700 Patch coming. 999257 1 228975 mark.lam 2014-04-09 12:44:05 -0700 Created attachment 228975 the patch 999258 2 228975 mhahnenberg 2014-04-09 12:46:48 -0700 Comment on attachment 228975 the patch View in context: https://bugs.webkit.org/attachment.cgi?id=228975&action=review > Source/JavaScriptCore/llint/LowLevelInterpreter64.asm:194 > + loadp ProtoCallFrame::paddedArgCount[protoCallFrame], temp3 Doesn't this change the behavior? We will end up loading an extra 32 bits. Was the old behavior a bug? 999259 3 mark.lam 2014-04-09 12:50:27 -0700 (In reply to comment #2) > (From update of attachment 228975 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=228975&action=review > > > Source/JavaScriptCore/llint/LowLevelInterpreter64.asm:194 > > + loadp ProtoCallFrame::paddedArgCount[protoCallFrame], temp3 > > Doesn't this change the behavior? We will end up loading an extra 32 bits. Was the old behavior a bug? Yes, the old behavior is a bug. Alternatively, we can change ProtoCallFrame::paddedArgCount to an unsigned instead if a size_t. 999260 4 228975 mhahnenberg 2014-04-09 12:53:06 -0700 Comment on attachment 228975 the patch View in context: https://bugs.webkit.org/attachment.cgi?id=228975&action=review >>> Source/JavaScriptCore/llint/LowLevelInterpreter64.asm:194 >>> + loadp ProtoCallFrame::paddedArgCount[protoCallFrame], temp3 >> >> Doesn't this change the behavior? We will end up loading an extra 32 bits. Was the old behavior a bug? > > Yes, the old behavior is a bug. Alternatively, we can change ProtoCallFrame::paddedArgCount to an unsigned instead if a size_t. Hmm, I'm unfamiliar with ProtoCallFrame. Does the size of the struct need to be consistent between 32-bit and 64-bit? If so then we should definitely change the type to uint32_t. 999261 5 228975 mhahnenberg 2014-04-09 13:03:03 -0700 Comment on attachment 228975 the patch View in context: https://bugs.webkit.org/attachment.cgi?id=228975&action=review >>>> Source/JavaScriptCore/llint/LowLevelInterpreter64.asm:194 >>>> + loadp ProtoCallFrame::paddedArgCount[protoCallFrame], temp3 >>> >>> Doesn't this change the behavior? We will end up loading an extra 32 bits. Was the old behavior a bug? >> >> Yes, the old behavior is a bug. Alternatively, we can change ProtoCallFrame::paddedArgCount to an unsigned instead if a size_t. > > Hmm, I'm unfamiliar with ProtoCallFrame. Does the size of the struct need to be consistent between 32-bit and 64-bit? If so then we should definitely change the type to uint32_t. Okay, looks like the ProtoCallFrame is the dummy call frame we use when we're about to start executing JS. I vote that we make ProtoCallFrame's arg count match the type of CallFrame, which looks like an int32_t. (Register::payload returns an int32_t, which then gets casted to a size_t inside CallFrame::argumentCountIncludingThis). So maybe uint32_t makes the most sense here. 999265 6 228978 mark.lam 2014-04-09 13:17:43 -0700 Created attachment 228978 revised patch. 999267 7 228978 mhahnenberg 2014-04-09 13:19:54 -0700 Comment on attachment 228978 revised patch. r=me! 999270 8 mark.lam 2014-04-09 13:24:39 -0700 Thanks. Landed in r167031: <http://trac.webkit.org/r167031>. 999277 9 ggaren 2014-04-09 14:02:28 -0700 Does this fix a bug? If so, it should come with a description of the bug and a regression test. 999280 10 mark.lam 2014-04-09 14:06:55 -0700 (In reply to comment #9) > Does this fix a bug? If so, it should come with a description of the bug and a regression test. The bug only manifests on a big endian system and was caught as soon as Tomas ran the JSC stress tests. I’ll try to create a more isolated test case and post that later. 999291 11 ggaren 2014-04-09 14:23:49 -0700 > The bug only manifests on a big endian system and was caught as soon as Tomas ran the JSC stress tests. I’ll try to create a more isolated test case and post that later. If this patch fixes a bug that's already covered by existing regression test(s), then you don't need to write a new regression test, and instead you should mention that the existing regression test(s) in your ChangeLog. 999292 12 mark.lam 2014-04-09 14:24:40 -0700 (In reply to comment #11) > > The bug only manifests on a big endian system and was caught as soon as Tomas ran the JSC stress tests. I’ll try to create a more isolated test case and post that later. > > If this patch fixes a bug that's already covered by existing regression test(s), then you don't need to write a new regression test, and instead you should mention that the existing regression test(s) in your ChangeLog. I will update the ChangeLog comment. 999308 13 mark.lam 2014-04-09 14:53:45 -0700 ChangeLog comment updated in r167037: <http://trac.webkit.org/changeset/167037> 228975 2014-04-09 12:44:05 -0700 2014-04-09 13:17:43 -0700 the patch bug-131449.patch text/plain 3164 mark.lam SW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9DaGFuZ2VMb2cKPT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291 cmNlL0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkocmV2aXNpb24gMTY3MDIxKQorKysgU291cmNl L0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDE5IEBA CisyMDE0LTA0LTA5ICBNYXJrIExhbSAgPG1hcmsubGFtQGFwcGxlLmNvbT4KKworICAgICAgICBF bnN1cmUgdGhhdCBMTElOVCBhY2Nlc3Npbmcgb2YgdGhlIFByb3RvQ2FsbEZyYW1lIGlzIGJpZyBl bmRpYW4gZnJpZW5kbHkuCisgICAgICAgIDxodHRwczovL3dlYmtpdC5vcmcvYi8xMzE0NDk+CisK KyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgUHJvdG9DYWxs RnJhbWU6OnBhZGRlZEFyZ0NvdW50IGlzIGEgc2l6ZV90LiAgV2Ugc2hvdWxkIGJlIGxvYWRpbmcg aXQgd2l0aCBhCisgICAgICAgIGxvYWRwIGluc3RlYWQgb2YgYSBsb2FkaS4KKworICAgICAgICBX ZSBzaG91bGQgYWRkIHRoZSBQYXlMb2FkT2Zmc2V0IHRvIFByb3RvQ2FsbEZyYW1lOjphcmdDb3Vu dEFuZENvZGVPcmlnaW5WYWx1ZQorICAgICAgICB3aGVuIGxvYWRpbmcgdGhlIGFyZ0NvdW50Lgor CisgICAgICAgICogbGxpbnQvTG93TGV2ZWxJbnRlcnByZXRlcjMyXzY0LmFzbToKKyAgICAgICAg KiBsbGludC9Mb3dMZXZlbEludGVycHJldGVyNjQuYXNtOgorCiAyMDE0LTA0LTA4ICBPbGl2ZXIg SHVudCAgPG9saXZlckBhcHBsZS5jb20+CiAKICAgICAgICAgUmV3cml0ZSBGdW5jdGlvbi5iaW5k IGFzIGEgYnVpbHRpbgpJbmRleDogU291cmNlL0phdmFTY3JpcHRDb3JlL2xsaW50L0xvd0xldmVs SW50ZXJwcmV0ZXIzMl82NC5hc20KPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291cmNlL0phdmFTY3JpcHRDb3Jl L2xsaW50L0xvd0xldmVsSW50ZXJwcmV0ZXIzMl82NC5hc20JKHJldmlzaW9uIDE2NzAyMCkKKysr IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9sbGludC9Mb3dMZXZlbEludGVycHJldGVyMzJfNjQuYXNt CSh3b3JraW5nIGNvcHkpCkBAIC0yNDQsNyArMjQ0LDcgQEAgbWFjcm8gZG9DYWxsVG9KYXZhU2Ny aXB0KG1ha2VDYWxsKQogICAgICAgICBsb2FkcCA0NFtzcCwgdGVtcDNdLCBwcm90b0NhbGxGcmFt ZQogICAgIGVuZAogCi0gICAgbG9hZGkgUHJvdG9DYWxsRnJhbWU6OnBhZGRlZEFyZ0NvdW50W3By b3RvQ2FsbEZyYW1lXSwgdGVtcDIKKyAgICBsb2FkcCBQcm90b0NhbGxGcmFtZTo6cGFkZGVkQXJn Q291bnRbcHJvdG9DYWxsRnJhbWVdLCB0ZW1wMgogICAgIGFkZHAgQ2FsbEZyYW1lSGVhZGVyU2xv dHMsIHRlbXAyLCB0ZW1wMgogICAgIGxzaGlmdHAgMywgdGVtcDIKICAgICBzdWJwIGNmciwgdGVt cDIsIHRlbXAxCkBAIC0yOTEsOSArMjkxLDkgQEAgbWFjcm8gZG9DYWxsVG9KYXZhU2NyaXB0KG1h a2VDYWxsKQogICAgIHN0b3JlaSB0ZW1wMywgUGF5bG9hZE9mZnNldCArIENvZGVCbG9ja1tzcCwg dGVtcDEsIDhdCiAgICAgYnRpbnogdGVtcDEsIC5jb3B5SGVhZGVyTG9vcAogCi0gICAgbG9hZGkg UHJvdG9DYWxsRnJhbWU6OmFyZ0NvdW50QW5kQ29kZU9yaWdpblZhbHVlW3Byb3RvQ2FsbEZyYW1l XSwgdGVtcDIKKyAgICBsb2FkaSBQYXlsb2FkT2Zmc2V0ICsgUHJvdG9DYWxsRnJhbWU6OmFyZ0Nv dW50QW5kQ29kZU9yaWdpblZhbHVlW3Byb3RvQ2FsbEZyYW1lXSwgdGVtcDIKICAgICBzdWJpIDEs IHRlbXAyCi0gICAgbG9hZGkgUHJvdG9DYWxsRnJhbWU6OnBhZGRlZEFyZ0NvdW50W3Byb3RvQ2Fs bEZyYW1lXSwgdGVtcDMKKyAgICBsb2FkcCBQcm90b0NhbGxGcmFtZTo6cGFkZGVkQXJnQ291bnRb cHJvdG9DYWxsRnJhbWVdLCB0ZW1wMwogICAgIHN1YmkgMSwgdGVtcDMKIAogICAgIGJpZXEgdGVt cDIsIHRlbXAzLCAuY29weUFyZ3MKSW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9sbGludC9M b3dMZXZlbEludGVycHJldGVyNjQuYXNtCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFNvdXJjZS9KYXZhU2NyaXB0 Q29yZS9sbGludC9Mb3dMZXZlbEludGVycHJldGVyNjQuYXNtCShyZXZpc2lvbiAxNjcwMjApCisr KyBTb3VyY2UvSmF2YVNjcmlwdENvcmUvbGxpbnQvTG93TGV2ZWxJbnRlcnByZXRlcjY0LmFzbQko d29ya2luZyBjb3B5KQpAQCAtMTQ5LDcgKzE0OSw3IEBAIG1hY3JvIGRvQ2FsbFRvSmF2YVNjcmlw dChtYWtlQ2FsbCkKICAgICBzdG9yZXAgcHJldmlvdXNQQywgUmV0dXJuUENbY2ZyXQogICAgIHN0 b3JlcCBwcmV2aW91c0NGUiwgQ2FsbGVyRnJhbWVbY2ZyXQogCi0gICAgbG9hZGkgUHJvdG9DYWxs RnJhbWU6OnBhZGRlZEFyZ0NvdW50W3Byb3RvQ2FsbEZyYW1lXSwgdGVtcDIKKyAgICBsb2FkcCBQ cm90b0NhbGxGcmFtZTo6cGFkZGVkQXJnQ291bnRbcHJvdG9DYWxsRnJhbWVdLCB0ZW1wMgogICAg IGFkZHAgQ2FsbEZyYW1lSGVhZGVyU2xvdHMsIHRlbXAyLCB0ZW1wMgogICAgIGxzaGlmdHAgMywg dGVtcDIKICAgICBzdWJwIGNmciwgdGVtcDIsIHRlbXAxCkBAIC0xODksOSArMTg5LDkgQEAgbWFj cm8gZG9DYWxsVG9KYXZhU2NyaXB0KG1ha2VDYWxsKQogICAgIHN0b3JlcSB0ZW1wMywgQ29kZUJs b2NrW3NwLCB0ZW1wMSwgOF0KICAgICBidGlueiB0ZW1wMSwgLmNvcHlIZWFkZXJMb29wCiAKLSAg ICBsb2FkaSBQcm90b0NhbGxGcmFtZTo6YXJnQ291bnRBbmRDb2RlT3JpZ2luVmFsdWVbcHJvdG9D YWxsRnJhbWVdLCB0ZW1wMgorICAgIGxvYWRpIFBheWxvYWRPZmZzZXQgKyBQcm90b0NhbGxGcmFt ZTo6YXJnQ291bnRBbmRDb2RlT3JpZ2luVmFsdWVbcHJvdG9DYWxsRnJhbWVdLCB0ZW1wMgogICAg IHN1YmkgMSwgdGVtcDIKLSAgICBsb2FkaSBQcm90b0NhbGxGcmFtZTo6cGFkZGVkQXJnQ291bnRb cHJvdG9DYWxsRnJhbWVdLCB0ZW1wMworICAgIGxvYWRwIFByb3RvQ2FsbEZyYW1lOjpwYWRkZWRB cmdDb3VudFtwcm90b0NhbGxGcmFtZV0sIHRlbXAzCiAgICAgc3ViaSAxLCB0ZW1wMwogCiAgICAg YmllcSB0ZW1wMiwgdGVtcDMsIC5jb3B5QXJncwo= 228978 2014-04-09 13:17:43 -0700 2014-04-09 13:19:54 -0700 revised patch. bug-131449.patch text/plain 3615 mark.lam SW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9DaGFuZ2VMb2cKPT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291 cmNlL0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkocmV2aXNpb24gMTY3MDI5KQorKysgU291cmNl L0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDIzIEBA CisyMDE0LTA0LTA5ICBNYXJrIExhbSAgPG1hcmsubGFtQGFwcGxlLmNvbT4KKworICAgICAgICBF bnN1cmUgdGhhdCBMTElOVCBhY2Nlc3Npbmcgb2YgdGhlIFByb3RvQ2FsbEZyYW1lIGlzIGJpZyBl bmRpYW4gZnJpZW5kbHkuCisgICAgICAgIDxodHRwczovL3dlYmtpdC5vcmcvYi8xMzE0NDk+CisK KyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgQ2hhbmdlIFBy b3RvQ2FsbEZyYW1lOjpwYWRkZWRBcmdDb3VudCB0byBiZSBvZiB0eXBlIHVpbnQzMl90LiAgVGhl IGFyZ0NvdW50CisgICAgICAgIHRoYXQgaXQgcGFkcyBpcyBvZiB0eXBlIGludCBhbnl3YXkuICBJ dCBkb2Vzbid0IG5lZWQgdG8gYmUgNjQgYml0LiAgVGhpcworICAgICAgICBhbHNvIG1ha2VzIGl0 IHdvcmsgd2l0aCB0aGUgTExJTlQgd2hpY2ggaXMgbG9hZGluZyBpdCB3aXRoIGEgbG9hZGkKKyAg ICAgICAgaW5zdHJ1Y3Rpb24uCisKKyAgICAgICAgV2Ugc2hvdWxkIGFkZCB0aGUgUGF5TG9hZE9m ZnNldCB0byBQcm90b0NhbGxGcmFtZTo6YXJnQ291bnRBbmRDb2RlT3JpZ2luVmFsdWUKKyAgICAg ICAgd2hlbiBsb2FkaW5nIHRoZSBhcmdDb3VudC4KKworICAgICAgICAqIGludGVycHJldGVyL1By b3RvQ2FsbEZyYW1lLmg6CisgICAgICAgIChKU0M6OlByb3RvQ2FsbEZyYW1lOjpzZXRQYWRkZWRB cmdDb3VudCk6CisgICAgICAgICogbGxpbnQvTG93TGV2ZWxJbnRlcnByZXRlcjMyXzY0LmFzbToK KyAgICAgICAgKiBsbGludC9Mb3dMZXZlbEludGVycHJldGVyNjQuYXNtOgorCiAyMDE0LTA0LTA4 ICBPbGl2ZXIgSHVudCAgPG9saXZlckBhcHBsZS5jb20+CiAKICAgICAgICAgUmV3cml0ZSBGdW5j dGlvbi5iaW5kIGFzIGEgYnVpbHRpbgpJbmRleDogU291cmNlL0phdmFTY3JpcHRDb3JlL2ludGVy cHJldGVyL1Byb3RvQ2FsbEZyYW1lLmgKPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291cmNlL0phdmFTY3JpcHRD b3JlL2ludGVycHJldGVyL1Byb3RvQ2FsbEZyYW1lLmgJKHJldmlzaW9uIDE2NzAyMCkKKysrIFNv dXJjZS9KYXZhU2NyaXB0Q29yZS9pbnRlcnByZXRlci9Qcm90b0NhbGxGcmFtZS5oCSh3b3JraW5n IGNvcHkpCkBAIC0zNiw3ICszNiw3IEBAIHN0cnVjdCBQcm90b0NhbGxGcmFtZSB7CiAgICAgUmVn aXN0ZXIgY2FsbGVlVmFsdWU7CiAgICAgUmVnaXN0ZXIgYXJnQ291bnRBbmRDb2RlT3JpZ2luVmFs dWU7CiAgICAgUmVnaXN0ZXIgdGhpc0FyZzsKLSAgICBzaXplX3QgcGFkZGVkQXJnQ291bnQ7Cisg ICAgdWludDMyX3QgcGFkZGVkQXJnQ291bnQ7CiAgICAgSlNWYWx1ZSAqYXJnczsKIAogICAgIHZv aWQgaW5pdChDb2RlQmxvY2sqLCBKU1Njb3BlKiwgSlNPYmplY3QqLCBKU1ZhbHVlLCBpbnQsIEpT VmFsdWUqIG90aGVyQXJncyA9IDApOwpAQCAtNTMsNyArNTMsNyBAQCBzdHJ1Y3QgUHJvdG9DYWxs RnJhbWUgewogICAgIGludCBhcmd1bWVudENvdW50SW5jbHVkaW5nVGhpcygpIGNvbnN0IHsgcmV0 dXJuIGFyZ0NvdW50QW5kQ29kZU9yaWdpblZhbHVlLnBheWxvYWQoKTsgfQogICAgIGludCBhcmd1 bWVudENvdW50KCkgY29uc3QgeyByZXR1cm4gYXJndW1lbnRDb3VudEluY2x1ZGluZ1RoaXMoKSAt IDE7IH0KICAgICB2b2lkIHNldEFyZ3VtZW50Q291bnRJbmNsdWRpbmdUaGlzKGludCBjb3VudCkg eyBhcmdDb3VudEFuZENvZGVPcmlnaW5WYWx1ZS5wYXlsb2FkKCkgPSBjb3VudDsgfQotICAgIHZv aWQgc2V0UGFkZGVkQXJnQ291bnQoc2l6ZV90IGFyZ0NvdW50KSB7IHBhZGRlZEFyZ0NvdW50ID0g YXJnQ291bnQ7IH0KKyAgICB2b2lkIHNldFBhZGRlZEFyZ0NvdW50KHVpbnQzMl90IGFyZ0NvdW50 KSB7IHBhZGRlZEFyZ0NvdW50ID0gYXJnQ291bnQ7IH0KIAogICAgIHZvaWQgY2xlYXJDdXJyZW50 VlBDKCkgeyBhcmdDb3VudEFuZENvZGVPcmlnaW5WYWx1ZS50YWcoKSA9IDA7IH0KICAgICAKSW5k ZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9sbGludC9Mb3dMZXZlbEludGVycHJldGVyMzJfNjQu YXNtCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT0KLS0tIFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9sbGludC9Mb3dMZXZlbElu dGVycHJldGVyMzJfNjQuYXNtCShyZXZpc2lvbiAxNjcwMjApCisrKyBTb3VyY2UvSmF2YVNjcmlw dENvcmUvbGxpbnQvTG93TGV2ZWxJbnRlcnByZXRlcjMyXzY0LmFzbQkod29ya2luZyBjb3B5KQpA QCAtMjkxLDcgKzI5MSw3IEBAIG1hY3JvIGRvQ2FsbFRvSmF2YVNjcmlwdChtYWtlQ2FsbCkKICAg ICBzdG9yZWkgdGVtcDMsIFBheWxvYWRPZmZzZXQgKyBDb2RlQmxvY2tbc3AsIHRlbXAxLCA4XQog ICAgIGJ0aW56IHRlbXAxLCAuY29weUhlYWRlckxvb3AKIAotICAgIGxvYWRpIFByb3RvQ2FsbEZy YW1lOjphcmdDb3VudEFuZENvZGVPcmlnaW5WYWx1ZVtwcm90b0NhbGxGcmFtZV0sIHRlbXAyCisg ICAgbG9hZGkgUGF5bG9hZE9mZnNldCArIFByb3RvQ2FsbEZyYW1lOjphcmdDb3VudEFuZENvZGVP cmlnaW5WYWx1ZVtwcm90b0NhbGxGcmFtZV0sIHRlbXAyCiAgICAgc3ViaSAxLCB0ZW1wMgogICAg IGxvYWRpIFByb3RvQ2FsbEZyYW1lOjpwYWRkZWRBcmdDb3VudFtwcm90b0NhbGxGcmFtZV0sIHRl bXAzCiAgICAgc3ViaSAxLCB0ZW1wMwpJbmRleDogU291cmNlL0phdmFTY3JpcHRDb3JlL2xsaW50 L0xvd0xldmVsSW50ZXJwcmV0ZXI2NC5hc20KPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291cmNlL0phdmFTY3Jp cHRDb3JlL2xsaW50L0xvd0xldmVsSW50ZXJwcmV0ZXI2NC5hc20JKHJldmlzaW9uIDE2NzAyMCkK KysrIFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9sbGludC9Mb3dMZXZlbEludGVycHJldGVyNjQuYXNt CSh3b3JraW5nIGNvcHkpCkBAIC0xODksNyArMTg5LDcgQEAgbWFjcm8gZG9DYWxsVG9KYXZhU2Ny aXB0KG1ha2VDYWxsKQogICAgIHN0b3JlcSB0ZW1wMywgQ29kZUJsb2NrW3NwLCB0ZW1wMSwgOF0K ICAgICBidGlueiB0ZW1wMSwgLmNvcHlIZWFkZXJMb29wCiAKLSAgICBsb2FkaSBQcm90b0NhbGxG cmFtZTo6YXJnQ291bnRBbmRDb2RlT3JpZ2luVmFsdWVbcHJvdG9DYWxsRnJhbWVdLCB0ZW1wMgor ICAgIGxvYWRpIFBheWxvYWRPZmZzZXQgKyBQcm90b0NhbGxGcmFtZTo6YXJnQ291bnRBbmRDb2Rl T3JpZ2luVmFsdWVbcHJvdG9DYWxsRnJhbWVdLCB0ZW1wMgogICAgIHN1YmkgMSwgdGVtcDIKICAg ICBsb2FkaSBQcm90b0NhbGxGcmFtZTo6cGFkZGVkQXJnQ291bnRbcHJvdG9DYWxsRnJhbWVdLCB0 ZW1wMwogICAgIHN1YmkgMSwgdGVtcDMK