157805 2016-05-17 11:48:28 -0700 REGRESSION(r192855): Math.random() always produces the same first 7 decimal points the first two invocations 2016-05-17 15:51:57 -0700 1 1 1 Unclassified WebKit JavaScriptCore WebKit Nightly Build Unspecified Unspecified RESOLVED FIXED InRadar P2 Normal --- 1 joepeck joepeck achristensen benjamin cdumez cmarcelo commit-queue fpizlo ggaren joepeck mark.lam sam webkit-bug-importer oldest_to_newest 1193897 0 joepeck 2016-05-17 11:48:28 -0700 * SUMMARY Math.random() always produces the same first 7 decimal points the first two invocations * STEPS TO REPRODUCE 1. Inspect about:blank 2. Run Math.random() twice => first 7 decimals are identical * EXAMPLES > Math.random() < 0.39641653952203815 < 0.39641654154667616 > Math.random() < 0.7208129766006761 < 0.7208129575360934 * NOTES Seen in a comment on Hacker News: <https://news.ycombinator.com/item?id=11711912> > 2) Math.random() returns the same result the first two calls on page load. 1193900 1 fpizlo 2016-05-17 11:50:06 -0700 Wow, that's hilarious. 1193906 2 joepeck 2016-05-17 12:01:18 -0700 Caused by r192855: <http://trac.webkit.org/changeset/192855> 1193908 3 ggaren 2016-05-17 12:01:46 -0700 Probably too many zeroes in the initial seed, since our seed API is 32 bit. You can probably fix this by: uint64_t seed64 = seed; m_low = (seed64 << 32) | seed64; m_high = (seed64 << 32) | seed64; Or you could update the seed API to be 128bit. 1193918 4 webkit-bug-importer 2016-05-17 12:21:48 -0700 <rdar://problem/26327851> 1193962 5 joepeck 2016-05-17 13:33:10 -0700 (In reply to comment #3) > Probably too many zeroes in the initial seed, since our seed API is 32 bit. > You can probably fix this by: > > uint64_t seed64 = seed; > m_low = (seed64 << 32) | seed64; > m_high = (seed64 << 32) | seed64; > After this the first 4 digits stayed nearly identical: >>> Math.random() 0.41647092963078913 >>> Math.random() 0.41646612834590113 >>> Math.random() 0.47840231815551537 >>> Math.random() 0.47839187513838377 How about we just advance once after setting the seed? m_low = seed; m_high = seed; advance(); Do you see any disadvantages to this? 1193987 6 ggaren 2016-05-17 14:29:59 -0700 > How about we just advance once after setting the seed? > > m_low = seed; > m_high = seed; > advance(); > > Do you see any disadvantages to this? LOL. Brain fart. It's guaranteed that no better solution is possible. We have 128 bits of state that we want to set based on 32 bits of input. All solutions to this problem include some kind of bias unless they are... random! 1193993 7 279167 joepeck 2016-05-17 14:47:44 -0700 Created attachment 279167 [PATCH] Proposed Fix 1193995 8 joepeck 2016-05-17 14:48:51 -0700 An alternative test approach could be TestWebKitAPI creating a WeakRandom, or a jsc API that allows setting the random seed. But this tests Math.random which is web exposed so probably the safest. 1194017 9 279167 ggaren 2016-05-17 15:29:31 -0700 Comment on attachment 279167 [PATCH] Proposed Fix EWS failure seems spurious. 1194027 10 279167 commit-queue 2016-05-17 15:51:49 -0700 Comment on attachment 279167 [PATCH] Proposed Fix Clearing flags on attachment: 279167 Committed r201053: <http://trac.webkit.org/changeset/201053> 1194028 11 commit-queue 2016-05-17 15:51:57 -0700 All reviewed patches have been landed. Closing bug. 279167 2016-05-17 14:47:44 -0700 2016-05-17 15:51:49 -0700 [PATCH] Proposed Fix random-1.patch text/plain 4707 joepeck ZGlmZiAtLWdpdCBhL0xheW91dFRlc3RzL0NoYW5nZUxvZyBiL0xheW91dFRlc3RzL0NoYW5nZUxv ZwppbmRleCBjYmEyZTMwLi43MjA4NzllIDEwMDY0NAotLS0gYS9MYXlvdXRUZXN0cy9DaGFuZ2VM b2cKKysrIGIvTGF5b3V0VGVzdHMvQ2hhbmdlTG9nCkBAIC0xLDMgKzEsMTggQEAKKzIwMTYtMDUt MTcgIEpvc2VwaCBQZWNvcmFybyAgPHBlY29yYXJvQGFwcGxlLmNvbT4KKworICAgICAgICBSRUdS RVNTSU9OKHIxOTI4NTUpOiBNYXRoLnJhbmRvbSgpIGFsd2F5cyBwcm9kdWNlcyB0aGUgc2FtZSBm aXJzdCA3IGRlY2ltYWwgcG9pbnRzIHRoZSBmaXJzdCB0d28gaW52b2NhdGlvbnMKKyAgICAgICAg aHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTE1NzgwNQorICAgICAgICA8 cmRhcjovL3Byb2JsZW0vMjYzMjc4NTE+CisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChP T1BTISkuCisKKyAgICAgICAgKiBqcy9kb20vbWF0aC1yYW5kb20taW5pdGlhbC12YWx1ZXMtZXhw ZWN0ZWQudHh0OiBBZGRlZC4KKyAgICAgICAgKiBqcy9kb20vbWF0aC1yYW5kb20taW5pdGlhbC12 YWx1ZXMuaHRtbDogQWRkZWQuCisgICAgICAgICoganMvcmVzb3VyY2VzL21hdGgtcmFuZG9tLWlu aXRpYWwtdmFsdWVzLWlmcmFtZS5odG1sOiBBZGRlZC4KKyAgICAgICAgVGVzdCB0aGF0IGxlc3Mg dGhlbiA1JSBvZiB0aGUgdGltZSwgZWFybHkgTWF0aC5yYW5kb20gaW52b2NhdGlvbnMKKyAgICAg ICAgcHJvZHVjZSB2ZXJ5IHNpbWlsaWFyIHZhbHVlcy4gQmVmb3JlIHRoaXMgY2hhbmdlIHdlIHdl cmUgZmFpbGluZworICAgICAgICAxMDAlLCBidXQgYWZ0ZXIgd2Ugc2VlIHNpbWlsaWFyIHZhbHVl cyBtb3N0bHkgYXJvdW5kIDAtMyUuCisKIDIwMTYtMDUtMTAgIEpvc2VwaCBQZWNvcmFybyAgPHBl Y29yYXJvQGFwcGxlLmNvbT4KIAogICAgICAgICBJbXByb3ZlIGVycm9yIG1lc3NhZ2VzIGZvciBh Y2Nlc3NpbmcgYXJndW1lbnRzLmNhbGxlZSBhbmQgc2ltaWxhciBnZXR0ZXJzIGluIHN0cmljdCBt b2RlCmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy9qcy9kb20vbWF0aC1yYW5kb20taW5pdGlhbC12 YWx1ZXMtZXhwZWN0ZWQudHh0IGIvTGF5b3V0VGVzdHMvanMvZG9tL21hdGgtcmFuZG9tLWluaXRp YWwtdmFsdWVzLWV4cGVjdGVkLnR4dApuZXcgZmlsZSBtb2RlIDEwMDY0NAppbmRleCAwMDAwMDAw Li43ZWFmYjc5Ci0tLSAvZGV2L251bGwKKysrIGIvTGF5b3V0VGVzdHMvanMvZG9tL21hdGgtcmFu ZG9tLWluaXRpYWwtdmFsdWVzLWV4cGVjdGVkLnR4dApAQCAtMCwwICsxLDEwIEBACitUaGlzIHRl c3RzIGVhcmx5IE1hdGgucmFuZG9tIHZhbHVlcyBhcmUgbm90IGNvbnNpc3RlbnRseSBuZWFybHkg aWRlbnRpY2FsLgorCitPbiBzdWNjZXNzLCB5b3Ugd2lsbCBzZWUgYSBzZXJpZXMgb2YgIlBBU1Mi IG1lc3NhZ2VzLCBmb2xsb3dlZCBieSAiVEVTVCBDT01QTEVURSIuCisKKworUEFTUyBmcmFtZXNX aXRoU2ltaWxpYXJFYXJseVJhbmRvbVZhbHVlcyA8PSA1IGlzIHRydWUKK1BBU1Mgc3VjY2Vzc2Z1 bGx5UGFyc2VkIGlzIHRydWUKKworVEVTVCBDT01QTEVURQorCmRpZmYgLS1naXQgYS9MYXlvdXRU ZXN0cy9qcy9kb20vbWF0aC1yYW5kb20taW5pdGlhbC12YWx1ZXMuaHRtbCBiL0xheW91dFRlc3Rz L2pzL2RvbS9tYXRoLXJhbmRvbS1pbml0aWFsLXZhbHVlcy5odG1sCm5ldyBmaWxlIG1vZGUgMTAw NjQ0CmluZGV4IDAwMDAwMDAuLjQxZGQ1ZjkKLS0tIC9kZXYvbnVsbAorKysgYi9MYXlvdXRUZXN0 cy9qcy9kb20vbWF0aC1yYW5kb20taW5pdGlhbC12YWx1ZXMuaHRtbApAQCAtMCwwICsxLDQ0IEBA Cis8IURPQ1RZUEUgSFRNTD4KKzxodG1sPgorPGhlYWQ+Cis8c2NyaXB0IHNyYz0iLi4vLi4vcmVz b3VyY2VzL2pzLXRlc3QtcHJlLmpzIj48L3NjcmlwdD4KKzwvaGVhZD4KKzxib2R5PgorPHNjcmlw dD4KK3dpbmRvdy5qc1Rlc3RJc0FzeW5jID0gdHJ1ZTsKKworZGVzY3JpcHRpb24oIlRoaXMgdGVz dHMgZWFybHkgTWF0aC5yYW5kb20gdmFsdWVzIGFyZSBub3QgY29uc2lzdGVudGx5IG5lYXJseSBp ZGVudGljYWwuIik7CisKK3ZhciBzZWVuID0gMDsKK3ZhciBleHBlY3RlZCA9IDEwMDsKK3ZhciBm cmFtZXNXaXRoU2ltaWxpYXJFYXJseVJhbmRvbVZhbHVlcyA9IDA7Cit2YXIgbGFzdFJhbmRvbTEg PSAwOworCit3aW5kb3cuYWRkRXZlbnRMaXN0ZW5lcigibWVzc2FnZSIsIGZ1bmN0aW9uKGV2ZW50 KSB7CisgICAgaWYgKGV2ZW50LmRhdGEuc2ltaWxpYXIpCisgICAgICAgIGZyYW1lc1dpdGhTaW1p bGlhckVhcmx5UmFuZG9tVmFsdWVzKys7CisKKyAgICBpZiAoZXZlbnQuZGF0YS5yYW5kb20xID09 PSBsYXN0UmFuZG9tMSkKKyAgICAgICAgdGVzdEZhaWxlZCgiUmFuZG9tIG51bWJlciBzaG91bGQg YmUgZGlmZmVyZW50IGJldHdlZW4gZnJhbWVzIik7CisgICAgbGFzdFJhbmRvbTEgPSBldmVudC5k YXRhLnJhbmRvbTE7CisKKyAgICBpZiAoKytzZWVuID09PSBleHBlY3RlZCkgeworICAgICAgICBz aG91bGRCZVRydWUoImZyYW1lc1dpdGhTaW1pbGlhckVhcmx5UmFuZG9tVmFsdWVzIDw9IDUiKTsK KyAgICAgICAgaWYgKGZyYW1lc1dpdGhTaW1pbGlhckVhcmx5UmFuZG9tVmFsdWVzID4gNSkKKyAg ICAgICAgICAgIGRlYnVnKCJmcmFtZXNXaXRoU2ltaWxpYXJFYXJseVJhbmRvbVZhbHVlczogIiAr IGZyYW1lc1dpdGhTaW1pbGlhckVhcmx5UmFuZG9tVmFsdWVzKTsKKyAgICAgICAgZmluaXNoSlNU ZXN0KCk7CisgICAgfQorfSk7CisKK2Z1bmN0aW9uIGNyZWF0ZUZyYW1lKCkgeworICAgIHZhciBp ZnJhbWUgPSBkb2N1bWVudC5jcmVhdGVFbGVtZW50KCJpZnJhbWUiKTsKKyAgICBpZnJhbWUuc3Jj ID0gIi4uL3Jlc291cmNlcy9tYXRoLXJhbmRvbS1pbml0aWFsLXZhbHVlcy1pZnJhbWUuaHRtbCI7 CisgICAgZG9jdW1lbnQuYm9keS5hcHBlbmRDaGlsZChpZnJhbWUpOworfQorCitmb3IgKHZhciBp ID0gMDsgaSA8IGV4cGVjdGVkOyArK2kpCisgICAgY3JlYXRlRnJhbWUoKTsKKzwvc2NyaXB0Pgor PHNjcmlwdCBzcmM9Ii4uLy4uL3Jlc291cmNlcy9qcy10ZXN0LXBvc3QuanMiPjwvc2NyaXB0Pgor PC9ib2R5PgorPC9odG1sPgpkaWZmIC0tZ2l0IGEvTGF5b3V0VGVzdHMvanMvcmVzb3VyY2VzL21h dGgtcmFuZG9tLWluaXRpYWwtdmFsdWVzLWlmcmFtZS5odG1sIGIvTGF5b3V0VGVzdHMvanMvcmVz b3VyY2VzL21hdGgtcmFuZG9tLWluaXRpYWwtdmFsdWVzLWlmcmFtZS5odG1sCm5ldyBmaWxlIG1v ZGUgMTAwNjQ0CmluZGV4IDAwMDAwMDAuLjVjZTE0YmUKLS0tIC9kZXYvbnVsbAorKysgYi9MYXlv dXRUZXN0cy9qcy9yZXNvdXJjZXMvbWF0aC1yYW5kb20taW5pdGlhbC12YWx1ZXMtaWZyYW1lLmh0 bWwKQEAgLTAsMCArMSw2IEBACis8c2NyaXB0PgordmFyIHJhbmRvbTEgPSBNYXRoLnJhbmRvbSgp OwordmFyIHJhbmRvbTIgPSBNYXRoLnJhbmRvbSgpOwordmFyIHNpbWlsaWFyID0gTWF0aC5mbG9v cihyYW5kb20xICogMTAwMCkgPT09IE1hdGguZmxvb3IocmFuZG9tMiAqIDEwMDApOword2luZG93 LnRvcC5wb3N0TWVzc2FnZSh7c2ltaWxpYXIsIHJhbmRvbTEsIHJhbmRvbTJ9LCAiKiIpOworPC9z Y3JpcHQ+CmRpZmYgLS1naXQgYS9Tb3VyY2UvV1RGL0NoYW5nZUxvZyBiL1NvdXJjZS9XVEYvQ2hh bmdlTG9nCmluZGV4IDUyZjY5ZDcuLjIxOTZmMTYgMTAwNjQ0Ci0tLSBhL1NvdXJjZS9XVEYvQ2hh bmdlTG9nCisrKyBiL1NvdXJjZS9XVEYvQ2hhbmdlTG9nCkBAIC0xLDMgKzEsMTYgQEAKKzIwMTYt MDUtMTcgIEpvc2VwaCBQZWNvcmFybyAgPHBlY29yYXJvQGFwcGxlLmNvbT4KKworICAgICAgICBS RUdSRVNTSU9OKHIxOTI4NTUpOiBNYXRoLnJhbmRvbSgpIGFsd2F5cyBwcm9kdWNlcyB0aGUgc2Ft ZSBmaXJzdCA3IGRlY2ltYWwgcG9pbnRzIHRoZSBmaXJzdCB0d28gaW52b2NhdGlvbnMKKyAgICAg ICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTE1NzgwNQorICAgICAg ICA8cmRhcjovL3Byb2JsZW0vMjYzMjc4NTE+CisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZ IChPT1BTISkuCisKKyAgICAgICAgKiB3dGYvV2Vha1JhbmRvbS5oOgorICAgICAgICAoV1RGOjpX ZWFrUmFuZG9tOjpzZXRTZWVkKToKKyAgICAgICAgQWR2YW5jZSBvbmNlIHRvIHJhbmRvbWl6ZSB0 aGUgMzJiaXQgc2VlZCBhY3Jvc3MgdGhlIDEyOGJpdCBzdGF0ZQorICAgICAgICBhbmQgYXZvaWQg cmUtdXNpbmcgNjRiaXRzIG9mIHN0YXRlIGluIHRoZSBzZWNvbmQgYWR2YW5jZS4KKwogMjAxNi0w NS0xNyAgWW91ZW5uIEZhYmxldCAgPHlvdWVubi5mYWJsZXRAY3JmLmNhbm9uLmZyPgogCiAgICAg ICAgIERPTVByb21pc2Ugc2hvdWxkIG9ubHkgcmVzdHJpY3QgdGhlIHJlc29sdXRpb24gdHlwZQpk aWZmIC0tZ2l0IGEvU291cmNlL1dURi93dGYvV2Vha1JhbmRvbS5oIGIvU291cmNlL1dURi93dGYv V2Vha1JhbmRvbS5oCmluZGV4IDljMGY1ODkuLjRkNTllNDQgMTAwNjQ0Ci0tLSBhL1NvdXJjZS9X VEYvd3RmL1dlYWtSYW5kb20uaAorKysgYi9Tb3VyY2UvV1RGL3d0Zi9XZWFrUmFuZG9tLmgKQEAg LTU2LDYgKzU2LDcgQEAgcHVibGljOgogCiAgICAgICAgIG1fbG93ID0gc2VlZDsKICAgICAgICAg bV9oaWdoID0gc2VlZDsKKyAgICAgICAgYWR2YW5jZSgpOwogICAgIH0KIAogICAgIHVuc2lnbmVk IHNlZWQoKSBjb25zdCB7IHJldHVybiBtX3NlZWQ7IH0K