99793 2012-10-18 18:49:38 -0700 make move_overwritten_baselines_to work again while rebaselining 2012-10-18 22:45:17 -0700 1 1 1 Unclassified WebKit New Bugs 528+ (Nightly build) Unspecified Unspecified RESOLVED FIXED P2 Normal --- 1 dpranke dpranke abarth dglazkov ojan rniwa tony webkit.review.bot zan oldest_to_newest 745874 0 dpranke 2012-10-18 18:49:38 -0700 make move_overwritten_baselines_to work again while rebaselining 745876 1 169530 dpranke 2012-10-18 18:51:32 -0700 Created attachment 169530 Patch 745881 2 dpranke 2012-10-18 18:58:26 -0700 It looks like this code broke at some point (possibly at least partially in r121706 / bug 90396, but probably partially in other patches as well). Since I need to bring up the new Chromium Mountain Lion baselines, it would be nice to make this work again in garden-o-matic. This patch, I believe restores the previous functionality, and I think we should land it as-is before trying something else. However, two ideas ... 1) It is kinda inconvenient to have to locally modify builders.js to take advantage of this functionality. I'm thinking it might make more sense to also support a command line flag to garden-o-matic, like "webkit-patch garden-o-matic --move-overwritten-baselines-to 'WebKit Mac10.8=chromium-mac-lion'" or something. I don't think the syntax needs to be terribly elegant. 2) Obviously better would be to just make garden-o-matic handle this rebaselining case better. Namely, when updating a baseline, we should check to see if this will potentially alter the results for any other builder that might use that baseline and, if so, and if not already also planning to rebaseline that builder, copy the overwritten result down. I think, since we rebaseline a test for all bots at once with rebaseline-json, this should be pretty straightforward, and I'm not sure that this strategy has any flaws. But obviously I'd want to do this in a separate patch . Thoughts? 745968 3 ojan 2012-10-18 22:02:28 -0700 I strongly prefer 2. But for unexpected failures, I think we decided on keeping the current behavior since the common case is that the test hasn't run on all the bots yet, but that they will share results (e.g. all the mac bots will have the same result). This minimized churn and waiting on bots to cycle for unexpected failures, but gets 100% correctness for expected failures or new bots. So, for new bots (e.g. the mtlion bot), I think we still want a way to mark it as doing this fallback behavior, just not needing to hard-code the fallback. 745973 4 169530 ojan 2012-10-18 22:18:38 -0700 Comment on attachment 169530 Patch View in context: https://bugs.webkit.org/attachment.cgi?id=169530&action=review > Tools/Scripts/webkitpy/layout_tests/port/builders.py:44 > +# FIXME: When do you need this? This is a workaround for ports that don't have full bot coverage, e.g. win-7sp1 actualy corresponds to platform/win, but there's no win-7sp1 bot, so we hack the win-7sp0 bot to put baselines in platform/win and qt-linux is th eonly qt bot on the webkit.org waterfall to run tests. I'm not sure at the moment, why mac-mountainlion needs this though. That might be a bug. It's a total hack. Not sure what else we could do. I suppose we could try to encourage ports to get better coverage, except platform/qt is like platform/chromium (i.e. there's no such thing as a plain qt bot). So they would need full coverage for every platform/qt-* directory before we could remove it for them. There's probably a better way to do all this, but it hasn't yet occurred to me. :) 745976 5 dpranke 2012-10-18 22:23:49 -0700 (In reply to comment #4) > (From update of attachment 169530 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=169530&action=review > > > Tools/Scripts/webkitpy/layout_tests/port/builders.py:44 > > +# FIXME: When do you need this? > > This is a workaround for ports that don't have full bot coverage, e.g. win-7sp1 actualy corresponds to platform/win, but there's no win-7sp1 bot, so we hack the win-7sp0 bot to put baselines in platform/win and qt-linux is th eonly qt bot on the webkit.org waterfall to run tests. I'm not sure at the moment, why mac-mountainlion needs this though. That might be a bug. > > It's a total hack. Not sure what else we could do. I suppose we could try to encourage ports to get better coverage, except platform/qt is like platform/chromium (i.e. there's no such thing as a plain qt bot). So they would need full coverage for every platform/qt-* directory before we could remove it for them. > > There's probably a better way to do all this, but it hasn't yet occurred to me. :) Okay, I'll add something about this to the comments. 745977 6 dpranke 2012-10-18 22:26:06 -0700 (In reply to comment #3) > I strongly prefer 2. But for unexpected failures, I think we decided on keeping the current behavior since the common case is that the test hasn't run on all the bots yet, but that they will share results (e.g. all the mac bots will have the same result). I think you've convinced me that we shouldn't move overwritten baselines by default for unexpected failures. I'm not sure that it's worth providing a way to choose to move overwritten baselines in the UI; it might be sufficient to never move unexpected and always move expected? > This minimized churn and waiting on bots to cycle for unexpected failures, but gets 100% correctness for expected failures or new bots. Is there a difference between expected failures and new bots here? >So, for new bots (e.g. the mtlion bot), I think we still want a way to mark it as doing this fallback behavior, just not needing to hard-code the fallback. I'm not sure what you mean by this sentence. 745980 7 dpranke 2012-10-18 22:32:27 -0700 Committed r131856: <http://trac.webkit.org/changeset/131856> 745982 8 ojan 2012-10-18 22:34:14 -0700 (In reply to comment #6) > (In reply to comment #3) > > I strongly prefer 2. But for unexpected failures, I think we decided on keeping the current behavior since the common case is that the test hasn't run on all the bots yet, but that they will share results (e.g. all the mac bots will have the same result). > > I think you've convinced me that we shouldn't move overwritten baselines by default for unexpected failures. I'm not sure that it's worth providing a way to choose to move overwritten baselines in the UI; it might be sufficient to never move unexpected and always move expected? The latter is what I'm picturing. > > This minimized churn and waiting on bots to cycle for unexpected failures, but gets 100% correctness for expected failures or new bots. > > Is there a difference between expected failures and new bots here? No. > >So, for new bots (e.g. the mtlion bot), I think we still want a way to mark it as doing this fallback behavior, just not needing to hard-code the fallback. > > I'm not sure what you mean by this sentence. All I meant is the above. We need some way to mark new bots as new so that we no know to treat their unexpected failures as expected failures. Oh, I suppose the way you have it now with a special TestExpectations file, they already all show up under expected failures, so there's no extra plumbing. That's seems better. 745985 9 dpranke 2012-10-18 22:38:30 -0700 (In reply to comment #8) > > >So, for new bots (e.g. the mtlion bot), I think we still want a way to mark it as doing this fallback behavior, just not needing to hard-code the fallback. > > > > I'm not sure what you mean by this sentence. > > All I meant is the above. We need some way to mark new bots as new so that we no know to treat their unexpected failures as expected failures. Oh, I suppose the way you have it now with a special TestExpectations file, they already all show up under expected failures, so there's no extra plumbing. That's seems better. Yeah, it's really easy to make unexpected failures expected at this point. I'm not even sure that the separate TestExpectations file is worth it assuming the tools will just work (although I suppose for non-Chromium ports a separate file is kinda the default anyway). 745991 10 ojan 2012-10-18 22:42:28 -0700 > Yeah, it's really easy to make unexpected failures expected at this point. I'm not even sure that the separate TestExpectations file is worth it assuming the tools will just work (although I suppose for non-Chromium ports a separate file is kinda the default anyway). The only problem with this is the current garden-o-matic UI makes it hard to work through just the list of MtLion expected failures. With unexpected failures it groups them in a more convenient way for that specific use-case. 745995 11 dpranke 2012-10-18 22:45:17 -0700 (In reply to comment #10) > > Yeah, it's really easy to make unexpected failures expected at this point. I'm not even sure that the separate TestExpectations file is worth it assuming the tools will just work (although I suppose for non-Chromium ports a separate file is kinda the default anyway). > > The only problem with this is the current garden-o-matic UI makes it hard to work through just the list of MtLion expected failures. With unexpected failures it groups them in a more convenient way for that specific use-case. Yeah, although at least this evening I was finding it really convenient to see the failures by name and just look to see if a given test was failing on ML or ML+other bots; the unexpected results page might not give you this context (although I need to look again). Fortunately, I still have a few thousand tests to go so I'll have plenty more opportunity to think about what might work better than what we have now :). 169530 2012-10-18 18:51:32 -0700 2012-10-18 22:18:37 -0700 Patch bug-99793-20121018185017.patch text/plain 13790 dpranke U3VidmVyc2lvbiBSZXZpc2lvbjogMTMxODI3CmRpZmYgLS1naXQgYS9Ub29scy9DaGFuZ2VMb2cg Yi9Ub29scy9DaGFuZ2VMb2cKaW5kZXggZGY0MTNhM2NmYjY3NGIxOTZkYzJlNjAyMjExNTdjZjll MGE0MjNiOS4uMTI2NDljNjU0ZGJmZWFiNWVmNThmOGQxYTA3ZTcyNzIyMzVlNjhhMCAxMDA2NDQK LS0tIGEvVG9vbHMvQ2hhbmdlTG9nCisrKyBiL1Rvb2xzL0NoYW5nZUxvZwpAQCAtMSw1ICsxLDM1 IEBACiAyMDEyLTEwLTE4ICBEaXJrIFByYW5rZSAgPGRwcmFua2VAY2hyb21pdW0ub3JnPgogCisg ICAgICAgIG1ha2UgbW92ZV9vdmVyd3JpdHRlbl9iYXNlbGluZXNfdG8gd29yayBhZ2FpbiB3aGls ZSByZWJhc2VsaW5pbmcKKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcu Y2dpP2lkPTk5NzkzCisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAg ICAgICAgV2UgdXNlZCB0byBoYXZlIGNvZGUgdGhhdCB3b3VsZCBoZWxwIHdoZW4gcmViYXNlbGlu aW5nIHJlc3VsdHMKKyAgICAgICAgZm9yIG5ldyBwb3J0cyAoZS5nLiwgd2hlbiBtb3ZpbmcgZnJv bSBMaW9uIHRvIE1vdW50YWluIExpb24pIGJ1dAorICAgICAgICBpdCBhcHBlYXJzIHRoYXQgY29k ZSBiaXRyb3R0ZWQgZHVyaW5nIGFsbCB0aGUgcmVmYWN0b3Jpbmcgc2luY2UKKyAgICAgICAgdGhl IGxhc3QgdGltZSB3ZSBuZWVkZWQgdGhpcy4gVGhpcyBwYXRjaCBtYWtlcyB0aGluZ3Mgd29yawor ICAgICAgICBhZ2FpbiwgYWRkcyBtb3JlIHRlc3RzIHRoYXQgd2lsbCBob3BlZnVsbHkgbWFrZSBj bGVhcmVyIHdoYXQKKyAgICAgICAgaXMgc3VwcG9zZWQgdG8gd29yaywgYW5kIHJlbmFtZXMgdmFy aW91cyBwYXJhbWV0ZXJzIHRvIGJlCisgICAgICAgIGNvbnNpc3RlbnQuCisKKyAgICAgICAgKiBT Y3JpcHRzL3dlYmtpdHB5L2xheW91dF90ZXN0cy9wb3J0L2J1aWxkZXJzLnB5OgorICAgICAgICAo bW92ZV9vdmVyd3JpdHRlbl9iYXNlbGluZXNfdG8pOgorICAgICAgICAoYnVpbGRlcl9wYXRoX2Zv cl9wb3J0X25hbWUpOgorICAgICAgICAqIFNjcmlwdHMvd2Via2l0cHkvdG9vbC9jb21tYW5kcy9y ZWJhc2VsaW5lLnB5OgorICAgICAgICAoUmViYXNlbGluZVRlc3QuX19pbml0X18pOgorICAgICAg ICAoUmViYXNlbGluZVRlc3QuX2NvcHlfZXhpc3RpbmdfYmFzZWxpbmUpOgorICAgICAgICAoUmVi YXNlbGluZVRlc3QuX3JlYmFzZWxpbmVfdGVzdCk6CisgICAgICAgIChSZWJhc2VsaW5lVGVzdC5l eGVjdXRlKToKKyAgICAgICAgKEFic3RyYWN0UGFyYWxsZWxSZWJhc2VsaW5lQ29tbWFuZC5fcmVi YXNlbGluZV9jb21tYW5kcyk6CisgICAgICAgICogU2NyaXB0cy93ZWJraXRweS90b29sL2NvbW1h bmRzL3JlYmFzZWxpbmVfdW5pdHRlc3QucHk6CisgICAgICAgICh0ZXN0X3JlYmFzZWxpbmVfanNv bl93aXRoX21vdmVfb3ZlcndyaXR0ZW5fYmFzZWxpbmVzX3RvKToKKyAgICAgICAgKHRlc3RfcmVi YXNlbGluZV90ZXN0X2ludGVybmFsX3dpdGhfbW92ZV9vdmVyd3JpdHRlbl9iYXNlbGluZXNfdG8p OgorICAgICAgICAqIFNjcmlwdHMvd2Via2l0cHkvdG9vbC9zZXJ2ZXJzL2dhcmRlbmluZ3NlcnZl cl91bml0dGVzdC5weToKKyAgICAgICAgKEdhcmRlbmluZ1NlcnZlclRlc3QudGVzdF9yZWJhc2Vs aW5lYWxsKToKKworMjAxMi0xMC0xOCAgRGlyayBQcmFua2UgIDxkcHJhbmtlQGNocm9taXVtLm9y Zz4KKwogICAgICAgICBbY2hyb21pdW1dIGdhcmRlbi1vLW1hdGljIHNob3VsZCBpZ25vcmUgdGhl IDEwLjggVGVzdHMgYm90CiAgICAgICAgIGh0dHBzOi8vYnVncy53ZWJraXQub3JnL3Nob3dfYnVn LmNnaT9pZD05OTc2NQogCmRpZmYgLS1naXQgYS9Ub29scy9TY3JpcHRzL3dlYmtpdHB5L2xheW91 dF90ZXN0cy9wb3J0L2J1aWxkZXJzLnB5IGIvVG9vbHMvU2NyaXB0cy93ZWJraXRweS9sYXlvdXRf dGVzdHMvcG9ydC9idWlsZGVycy5weQppbmRleCA0MTc1MzM5NzdhZThhNmJlNzc5ZGE0NmVkYWI0 NzQyMGE4MDVlNTIzLi5lNDFmMzQ3MjkxNDljMjJkYmNlYTBhODkwMDY2NDgyMzk0MTYwYmViIDEw MDY0NAotLS0gYS9Ub29scy9TY3JpcHRzL3dlYmtpdHB5L2xheW91dF90ZXN0cy9wb3J0L2J1aWxk ZXJzLnB5CisrKyBiL1Rvb2xzL1NjcmlwdHMvd2Via2l0cHkvbGF5b3V0X3Rlc3RzL3BvcnQvYnVp bGRlcnMucHkKQEAgLTM1LDYgKzM1LDE0IEBAIGZyb20gd2Via2l0cHkuY29tbW9uLm1lbW9pemVk IGltcG9ydCBtZW1vaXplZAogIyBJbiB0aGlzIGRpY3Rpb25hcnksIGVhY2ggaXRlbSBzdG9yZXM6 CiAjICogcG9ydF9uYW1lIC0tIGEgZnVsbHkgcXVhbGlmaWVkIHBvcnQgbmFtZQogIyAqIHNwZWNp ZmllcnMgLS0gYSBzZXQgb2Ygc3BlY2lmaWVycywgcmVwcmVzZW50aW5nIGNvbmZpZ3VyYXRpb25z IGNvdmVyZWQgYnkgdGhpcyBidWlsZGVyLgorIyAqIG1vdmVfb3ZlcndyaXR0ZW5fYmFzZWxpbmVz X3RvIC0tIChvcHRpb25hbCkgbGlzdCBvZiBwbGF0Zm9ybSBkaXJlY3RvcmllcyB0aGF0IHdlIHdp bGwgY29weSBhbiBleGlzdGluZworIyAgICAgIGJhc2VsaW5lIHRvIGJlZm9yZSBwdWxsaW5nIGRv d24gYSBuZXcgYmFzZWxpbmUgZHVyaW5nIHJlYmFzZWxpbmluZy4gVGhpcyBpcyB1c2VmdWwKKyMg ICAgICBmb3IgYnJpbmdpbmcgdXAgYSBuZXcgcG9ydCwgZm9yIGV4YW1wbGUgd2hlbiBhZGRpbmcg YSBMaW9uIHdhcyB0aGUgbW9zdCByZWNlbnQgTWFjIHZlcnNpb24gYW5kCisjICAgICAgd2Ugd2Fu dGVkIHRvIGJyaW5nIHVwIE1vdW50YWluIExpb24sIHdlIHdvdWxkIHdhbnQgdG8gY29weSBhbiBl eGlzdGluZyBiYXNlbGluZSBpbiBwbGF0Zm9ybS9tYWMKKyMgICAgICB0byBwbGF0Zm9ybS9tYWMt bW91bnRhaW5saW9uIGJlZm9yZSB1cGRhdGluZyB0aGUgcGxhdGZvcm0vbWFjIGVudHJ5LgorIyAq IHJlYmFzZWxpbmVfb3ZlcnJpZGVfZGlyIC0tIChvcHRpb25hbCkgZGlyZWN0b3J5IHRvIHB1dCBi YXNlbGluZXMgaW4gaW5zdGVhZCBvZiB3aGVyZSB5b3Ugd291bGQgbm9ybWFsbHkgcHV0IHRoZW0u CisjICAgICAgRklYTUU6IFdoZW4gZG8geW91IG5lZWQgdGhpcz8KKwogX2V4YWN0X21hdGNoZXMg PSB7CiAgICAgIyBUaGVzZSBidWlsZGVycyBhcmUgb24gYnVpbGQuY2hyb21pdW0ub3JnLgogICAg ICJXZWJLaXQgWFAiOiB7InBvcnRfbmFtZSI6ICJjaHJvbWl1bS13aW4teHAiLCAic3BlY2lmaWVy cyI6IHNldChbInhwIiwgInJlbGVhc2UiXSl9LApAQCAtMTE4LDYgKzEyNiwxMCBAQCBkZWYgcmVi YXNlbGluZV9vdmVycmlkZV9kaXIoYnVpbGRlcl9uYW1lKToKICAgICByZXR1cm4gX2V4YWN0X21h dGNoZXNbYnVpbGRlcl9uYW1lXS5nZXQoInJlYmFzZWxpbmVfb3ZlcnJpZGVfZGlyIiwgTm9uZSkK IAogCitkZWYgbW92ZV9vdmVyd3JpdHRlbl9iYXNlbGluZXNfdG8oYnVpbGRlcl9uYW1lKToKKyAg ICByZXR1cm4gX2V4YWN0X21hdGNoZXNbYnVpbGRlcl9uYW1lXS5nZXQoIm1vdmVfb3ZlcndyaXR0 ZW5fYmFzZWxpbmVzX3RvIiwgW10pCisKKwogZGVmIHBvcnRfbmFtZV9mb3JfYnVpbGRlcl9uYW1l KGJ1aWxkZXJfbmFtZSk6CiAgICAgaWYgYnVpbGRlcl9uYW1lIGluIF9leGFjdF9tYXRjaGVzOgog ICAgICAgICByZXR1cm4gX2V4YWN0X21hdGNoZXNbYnVpbGRlcl9uYW1lXVsicG9ydF9uYW1lIl0K QEAgLTEzNiw3ICsxNDgsMyBAQCBkZWYgYnVpbGRlcl9uYW1lX2Zvcl9wb3J0X25hbWUodGFyZ2V0 X3BvcnRfbmFtZSk6CiAKIGRlZiBidWlsZGVyX3BhdGhfZm9yX3BvcnRfbmFtZShwb3J0X25hbWUp OgogICAgIGJ1aWxkZXJfcGF0aF9mcm9tX25hbWUoYnVpbGRlcl9uYW1lX2Zvcl9wb3J0X25hbWUo cG9ydF9uYW1lKSkKLQotCi1kZWYgZmFsbGJhY2tfcG9ydF9uYW1lc19mb3JfbmV3X3BvcnQoYnVp bGRlcl9uYW1lKToKLSAgICByZXR1cm4gX2V4YWN0X21hdGNoZXNbYnVpbGRlcl9uYW1lXS5nZXQo Im1vdmVfb3ZlcndyaXR0ZW5fYmFzZWxpbmVzX3RvIiwgW10pCmRpZmYgLS1naXQgYS9Ub29scy9T Y3JpcHRzL3dlYmtpdHB5L3Rvb2wvY29tbWFuZHMvcmViYXNlbGluZS5weSBiL1Rvb2xzL1Njcmlw dHMvd2Via2l0cHkvdG9vbC9jb21tYW5kcy9yZWJhc2VsaW5lLnB5CmluZGV4IGNiOGZhMDAxZjQy YWVlZWUzYTc5MjJjMGFmM2NjNzI4MTI2MWNiYmQuLjg1OTk2MzI2MWNkMWQyMjBmMDI2NjBkMmNi YmNjODBmNThkYzAxMTYgMTAwNjQ0Ci0tLSBhL1Rvb2xzL1NjcmlwdHMvd2Via2l0cHkvdG9vbC9j b21tYW5kcy9yZWJhc2VsaW5lLnB5CisrKyBiL1Rvb2xzL1NjcmlwdHMvd2Via2l0cHkvdG9vbC9j b21tYW5kcy9yZWJhc2VsaW5lLnB5CkBAIC03NCw3ICs3NCw4IEBAIGNsYXNzIFJlYmFzZWxpbmVU ZXN0KEFic3RyYWN0UmViYXNlbGluaW5nQ29tbWFuZCk6CiAgICAgZGVmIF9faW5pdF9fKHNlbGYp OgogICAgICAgICBvcHRpb25zID0gWwogICAgICAgICAgICAgb3B0cGFyc2UubWFrZV9vcHRpb24o Ii0tYnVpbGRlciIsIGhlbHA9IkJ1aWxkZXIgdG8gcHVsbCBuZXcgYmFzZWxpbmVzIGZyb20iKSwK LSAgICAgICAgICAgIG9wdHBhcnNlLm1ha2Vfb3B0aW9uKCItLXBsYXRmb3JtLXRvLW1vdmUtdG8i LCBoZWxwPSJQbGF0Zm9ybSB0byBtb3ZlIGV4aXN0aW5nIGJhc2VsaW5lcyB0byBiZWZvcmUgcmVi YXNlbGluaW5nLiBUaGlzIGlzIGZvciBkZWFsaW5nIHdpdGggYnJpbmdpbmcgdXAgbmV3IHBvcnRz IHRoYXQgaW50ZXJhY3Qgd2l0aCBub24tdHJlZSBwb3J0aW9ucyBvZiB0aGUgZmFsbGJhY2sgZ3Jh cGguIiksCisgICAgICAgICAgICBvcHRwYXJzZS5tYWtlX29wdGlvbigiLS1tb3ZlLW92ZXJ3cml0 dGVuLWJhc2VsaW5lcy10byIsIGFjdGlvbj0iYXBwZW5kIiwgZGVmYXVsdD1bXSwKKyAgICAgICAg ICAgICAgICBoZWxwPSJQbGF0Zm9ybSB0byBtb3ZlIGV4aXN0aW5nIGJhc2VsaW5lcyB0byBiZWZv cmUgcmViYXNlbGluaW5nLiBUaGlzIGlzIGZvciBicmluZ2luZyB1cCBuZXcgcG9ydHMuIiksCiAg ICAgICAgICAgICBvcHRwYXJzZS5tYWtlX29wdGlvbigiLS10ZXN0IiwgaGVscD0iVGVzdCB0byBy ZWJhc2VsaW5lIiksCiAgICAgICAgIF0KICAgICAgICAgQWJzdHJhY3RSZWJhc2VsaW5pbmdDb21t YW5kLl9faW5pdF9fKHNlbGYsIG9wdGlvbnM9b3B0aW9ucykKQEAgLTkwLDEzICs5MSwxMyBAQCBj bGFzcyBSZWJhc2VsaW5lVGVzdChBYnN0cmFjdFJlYmFzZWxpbmluZ0NvbW1hbmQpOgogICAgICAg ICAgICAgcmV0dXJuIHNlbGYuX3Rvb2wuZmlsZXN5c3RlbS5qb2luKHBvcnQubGF5b3V0X3Rlc3Rz X2RpcigpLCAncGxhdGZvcm0nLCBvdmVycmlkZV9kaXIpCiAgICAgICAgIHJldHVybiBwb3J0LmJh c2VsaW5lX3ZlcnNpb25fZGlyKCkKIAotICAgIGRlZiBfY29weV9leGlzdGluZ19iYXNlbGluZShz ZWxmLCBwbGF0Zm9ybXNfdG9fbW92ZV9leGlzdGluZ19iYXNlbGluZXNfdG8sIHRlc3RfbmFtZSwg c3VmZml4KToKKyAgICBkZWYgX2NvcHlfZXhpc3RpbmdfYmFzZWxpbmUoc2VsZiwgbW92ZV9vdmVy d3JpdHRlbl9iYXNlbGluZXNfdG8sIHRlc3RfbmFtZSwgc3VmZml4KToKICAgICAgICAgb2xkX2Jh c2VsaW5lcyA9IFtdCiAgICAgICAgIG5ld19iYXNlbGluZXMgPSBbXQogCiAgICAgICAgICMgTmVl ZCB0byBnYXRoZXIgYWxsIHRoZSBiYXNlbGluZSBwYXRocyBiZWZvcmUgbW9kaWZ5aW5nIHRoZSBm aWxlc3lzdGVtIHNpbmNlCiAgICAgICAgICMgdGhlIG1vZGlmaWNhdGlvbnMgY2FuIGFmZmVjdCB0 aGUgcmVzdWx0cyBvZiBwb3J0LmV4cGVjdGVkX2ZpbGVuYW1lLgotICAgICAgICBmb3IgcGxhdGZv cm0gaW4gcGxhdGZvcm1zX3RvX21vdmVfZXhpc3RpbmdfYmFzZWxpbmVzX3RvOgorICAgICAgICBm b3IgcGxhdGZvcm0gaW4gbW92ZV9vdmVyd3JpdHRlbl9iYXNlbGluZXNfdG86CiAgICAgICAgICAg ICBwb3J0ID0gc2VsZi5fdG9vbC5wb3J0X2ZhY3RvcnkuZ2V0KHBsYXRmb3JtKQogICAgICAgICAg ICAgb2xkX2Jhc2VsaW5lID0gcG9ydC5leHBlY3RlZF9maWxlbmFtZSh0ZXN0X25hbWUsICIuIiAr IHN1ZmZpeCkKICAgICAgICAgICAgIGlmIG5vdCBzZWxmLl90b29sLmZpbGVzeXN0ZW0uZXhpc3Rz KG9sZF9iYXNlbGluZSk6CkBAIC0xNTIsMTUgKzE1MywxNSBAQCBjbGFzcyBSZWJhc2VsaW5lVGVz dChBYnN0cmFjdFJlYmFzZWxpbmluZ0NvbW1hbmQpOgogICAgIGRlZiBfZmlsZV9uYW1lX2Zvcl9l eHBlY3RlZF9yZXN1bHQoc2VsZiwgdGVzdF9uYW1lLCBzdWZmaXgpOgogICAgICAgICByZXR1cm4g IiVzLWV4cGVjdGVkLiVzIiAlIChzZWxmLl90ZXN0X3Jvb3QodGVzdF9uYW1lKSwgc3VmZml4KQog Ci0gICAgZGVmIF9yZWJhc2VsaW5lX3Rlc3Qoc2VsZiwgYnVpbGRlcl9uYW1lLCB0ZXN0X25hbWUs IHBsYXRmb3Jtc190b19tb3ZlX2V4aXN0aW5nX2Jhc2VsaW5lc190bywgc3VmZml4KToKKyAgICBk ZWYgX3JlYmFzZWxpbmVfdGVzdChzZWxmLCBidWlsZGVyX25hbWUsIHRlc3RfbmFtZSwgbW92ZV9v dmVyd3JpdHRlbl9iYXNlbGluZXNfdG8sIHN1ZmZpeCk6CiAgICAgICAgIHJlc3VsdHNfdXJsID0g c2VsZi5fcmVzdWx0c191cmwoYnVpbGRlcl9uYW1lKQogICAgICAgICBiYXNlbGluZV9kaXJlY3Rv cnkgPSBzZWxmLl9iYXNlbGluZV9kaXJlY3RvcnkoYnVpbGRlcl9uYW1lKQogCiAgICAgICAgIHNv dXJjZV9iYXNlbGluZSA9ICIlcy8lcyIgJSAocmVzdWx0c191cmwsIHNlbGYuX2ZpbGVfbmFtZV9m b3JfYWN0dWFsX3Jlc3VsdCh0ZXN0X25hbWUsIHN1ZmZpeCkpCiAgICAgICAgIHRhcmdldF9iYXNl bGluZSA9IHNlbGYuX3Rvb2wuZmlsZXN5c3RlbS5qb2luKGJhc2VsaW5lX2RpcmVjdG9yeSwgc2Vs Zi5fZmlsZV9uYW1lX2Zvcl9leHBlY3RlZF9yZXN1bHQodGVzdF9uYW1lLCBzdWZmaXgpKQogCi0g ICAgICAgIGlmIHBsYXRmb3Jtc190b19tb3ZlX2V4aXN0aW5nX2Jhc2VsaW5lc190bzoKLSAgICAg ICAgICAgIHNlbGYuX2NvcHlfZXhpc3RpbmdfYmFzZWxpbmUocGxhdGZvcm1zX3RvX21vdmVfZXhp c3RpbmdfYmFzZWxpbmVzX3RvLCB0ZXN0X25hbWUsIHN1ZmZpeCkKKyAgICAgICAgaWYgbW92ZV9v dmVyd3JpdHRlbl9iYXNlbGluZXNfdG86CisgICAgICAgICAgICBzZWxmLl9jb3B5X2V4aXN0aW5n X2Jhc2VsaW5lKG1vdmVfb3ZlcndyaXR0ZW5fYmFzZWxpbmVzX3RvLCB0ZXN0X25hbWUsIHN1ZmZp eCkKIAogICAgICAgICBfbG9nLmluZm8oIlJldHJpZXZpbmcgJXMuIiAlIHNvdXJjZV9iYXNlbGlu ZSkKICAgICAgICAgc2VsZi5fc2F2ZV9iYXNlbGluZShzZWxmLl90b29sLndlYi5nZXRfYmluYXJ5 KHNvdXJjZV9iYXNlbGluZSwgY29udmVydF80MDRfdG9fTm9uZT1UcnVlKSwgdGFyZ2V0X2Jhc2Vs aW5lKQpAQCAtMTcyLDcgKzE3Myw3IEBAIGNsYXNzIFJlYmFzZWxpbmVUZXN0KEFic3RyYWN0UmVi YXNlbGluaW5nQ29tbWFuZCk6CiAKICAgICBkZWYgZXhlY3V0ZShzZWxmLCBvcHRpb25zLCBhcmdz LCB0b29sKToKICAgICAgICAgc2VsZi5fYmFzZWxpbmVfc3VmZml4X2xpc3QgPSBvcHRpb25zLnN1 ZmZpeGVzLnNwbGl0KCcsJykKLSAgICAgICAgc2VsZi5fcmViYXNlbGluZV90ZXN0X2FuZF91cGRh dGVfZXhwZWN0YXRpb25zKG9wdGlvbnMuYnVpbGRlciwgb3B0aW9ucy50ZXN0LCBvcHRpb25zLnBs YXRmb3JtX3RvX21vdmVfdG8pCisgICAgICAgIHNlbGYuX3JlYmFzZWxpbmVfdGVzdF9hbmRfdXBk YXRlX2V4cGVjdGF0aW9ucyhvcHRpb25zLmJ1aWxkZXIsIG9wdGlvbnMudGVzdCwgb3B0aW9ucy5t b3ZlX292ZXJ3cml0dGVuX2Jhc2VsaW5lc190bykKICAgICAgICAgcHJpbnQganNvbi5kdW1wcyhz ZWxmLl9zY21fY2hhbmdlcykKIAogCkBAIC0yNjksNiArMjcwLDkgQEAgY2xhc3MgQWJzdHJhY3RQ YXJhbGxlbFJlYmFzZWxpbmVDb21tYW5kKEFic3RyYWN0RGVjbGFyYXRpdmVDb21tYW5kKToKICAg ICAgICAgICAgIGZvciBidWlsZGVyIGluIHNlbGYuX2J1aWxkZXJzX3RvX2ZldGNoX2Zyb20odGVz dF9saXN0W3Rlc3RdKToKICAgICAgICAgICAgICAgICBzdWZmaXhlcyA9ICcsJy5qb2luKHRlc3Rf bGlzdFt0ZXN0XVtidWlsZGVyXSkKICAgICAgICAgICAgICAgICBjbWRfbGluZSA9IFtwYXRoX3Rv X3dlYmtpdF9wYXRjaCwgJ3JlYmFzZWxpbmUtdGVzdC1pbnRlcm5hbCcsICctLXN1ZmZpeGVzJywg c3VmZml4ZXMsICctLWJ1aWxkZXInLCBidWlsZGVyLCAnLS10ZXN0JywgdGVzdF0KKyAgICAgICAg ICAgICAgICBtb3ZlX292ZXJ3cml0dGVuX2Jhc2VsaW5lc190byA9IGJ1aWxkZXJzLm1vdmVfb3Zl cndyaXR0ZW5fYmFzZWxpbmVzX3RvKGJ1aWxkZXIpCisgICAgICAgICAgICAgICAgZm9yIHBsYXRm b3JtIGluIG1vdmVfb3ZlcndyaXR0ZW5fYmFzZWxpbmVzX3RvOgorICAgICAgICAgICAgICAgICAg ICBjbWRfbGluZS5leHRlbmQoWyctLW1vdmUtb3ZlcndyaXR0ZW4tYmFzZWxpbmVzLXRvJywgcGxh dGZvcm1dKQogICAgICAgICAgICAgICAgIGNvbW1hbmRzLmFwcGVuZCh0dXBsZShbY21kX2xpbmUs IGN3ZF0pKQogICAgICAgICByZXR1cm4gY29tbWFuZHMKIApkaWZmIC0tZ2l0IGEvVG9vbHMvU2Ny aXB0cy93ZWJraXRweS90b29sL2NvbW1hbmRzL3JlYmFzZWxpbmVfdW5pdHRlc3QucHkgYi9Ub29s cy9TY3JpcHRzL3dlYmtpdHB5L3Rvb2wvY29tbWFuZHMvcmViYXNlbGluZV91bml0dGVzdC5weQpp bmRleCA2OGYxOWY2ZmFjODE4NmUwMGI1ZmJkZGI4YTZlZjVlOGYyMzdmMTAzLi4zNTM5NDI0NWZj N2RiZmI0ZDI0NDdkNTNkNDljNzI3NTVjZDRhYmJmIDEwMDY0NAotLS0gYS9Ub29scy9TY3JpcHRz L3dlYmtpdHB5L3Rvb2wvY29tbWFuZHMvcmViYXNlbGluZV91bml0dGVzdC5weQorKysgYi9Ub29s cy9TY3JpcHRzL3dlYmtpdHB5L3Rvb2wvY29tbWFuZHMvcmViYXNlbGluZV91bml0dGVzdC5weQpA QCAtNDU2LDMgKzQ1Niw2MCBAQCBNT0NLIHJ1bl9jb21tYW5kOiBbJ2VjaG8nLCAnb3B0aW1pemUt YmFzZWxpbmVzJywgJy0tc3VmZml4ZXMnLCAnd2F2LHR4dCxwbmcnLCAnbQogCiAgICAgICAgIGZp bmFsbHk6CiAgICAgICAgICAgICBidWlsZGVycy5fZXhhY3RfbWF0Y2hlcyA9IG9sZF9leGFjdF9t YXRjaGVzCisKKyAgICBkZWYgdGVzdF9yZWJhc2VsaW5lX2pzb25fd2l0aF9tb3ZlX292ZXJ3cml0 dGVuX2Jhc2VsaW5lc190byhzZWxmKToKKyAgICAgICAgb2xkX2V4YWN0X21hdGNoZXMgPSBidWls ZGVycy5fZXhhY3RfbWF0Y2hlcworICAgICAgICB0cnk6CisgICAgICAgICAgICBidWlsZGVycy5f ZXhhY3RfbWF0Y2hlcyA9IHsKKyAgICAgICAgICAgICAgICAiTU9DSyBidWlsZGVyIjogeyJwb3J0 X25hbWUiOiAidGVzdC1tYWMtbGVvcGFyZCIsICJzcGVjaWZpZXJzIjogc2V0KFsibW9jay1zcGVj aWZpZXIiXSl9LAorICAgICAgICAgICAgICAgICJNT0NLIGJ1aWxkZXIyIjogeyJwb3J0X25hbWUi OiAidGVzdC1tYWMtc25vd2xlb3BhcmQiLCAic3BlY2lmaWVycyI6IHNldChbIm1vY2stc3BlY2lm aWVyMiJdKSwKKyAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAibW92ZV9vdmVyd3Jp dHRlbl9iYXNlbGluZXNfdG8iOiBbInRlc3QtbWFjLWxlb3BhcmQiXX0sCisgICAgICAgICAgICB9 CisKKyAgICAgICAgICAgIGNvbW1hbmQgPSBSZWJhc2VsaW5lKCkKKyAgICAgICAgICAgIHRvb2wg PSBNb2NrVG9vbCgpCisgICAgICAgICAgICB0b29sLmV4ZWN1dGl2ZSA9IE1vY2tFeGVjdXRpdmUo c2hvdWxkX2xvZz1UcnVlKQorICAgICAgICAgICAgY29tbWFuZC5iaW5kX3RvX3Rvb2wodG9vbCkK KworICAgICAgICAgICAgZXhwZWN0ZWRfc3Rkb3V0ID0gIiIicmViYXNlbGluZS1qc29uOiB7J21v Y2svcGF0aC90by90ZXN0Lmh0bWwnOiB7J01PQ0sgYnVpbGRlcjInOiBbJ3R4dCcsICdwbmcnXX19 XG4iIiIKKyAgICAgICAgICAgIGV4cGVjdGVkX3N0ZGVyciA9ICIiIk1PQ0sgcnVuX2NvbW1hbmQ6 IFsnZWNobycsICdyZWJhc2VsaW5lLXRlc3QtaW50ZXJuYWwnLCAnLS1zdWZmaXhlcycsICd0eHQs cG5nJywgJy0tYnVpbGRlcicsICdNT0NLIGJ1aWxkZXIyJywgJy0tdGVzdCcsICdtb2NrL3BhdGgv dG8vdGVzdC5odG1sJywgJy0tbW92ZS1vdmVyd3JpdHRlbi1iYXNlbGluZXMtdG8nLCAndGVzdC1t YWMtbGVvcGFyZCddLCBjd2Q9L21vY2stY2hlY2tvdXQKK01PQ0sgcnVuX2NvbW1hbmQ6IFsnZWNo bycsICdvcHRpbWl6ZS1iYXNlbGluZXMnLCAnLS1zdWZmaXhlcycsICd0eHQscG5nJywgJ21vY2sv cGF0aC90by90ZXN0Lmh0bWwnXSwgY3dkPS9tb2NrLWNoZWNrb3V0CisiIiIKKworICAgICAgICAg ICAgb3B0aW9ucyA9IE1vY2tPcHRpb25zKG9wdGltaXplPVRydWUsIGJ1aWxkZXJzPVsiTU9DSyBi dWlsZGVyMiJdLCBzdWZmaXhlcz1bInR4dCxwbmciXSwgdmVyYm9zZT1UcnVlKQorICAgICAgICAg ICAgT3V0cHV0Q2FwdHVyZSgpLmFzc2VydF9vdXRwdXRzKHNlbGYsIGNvbW1hbmQuZXhlY3V0ZSwg W29wdGlvbnMsIFsibW9jay9wYXRoL3RvL3Rlc3QuaHRtbCJdLCB0b29sXSwKKyAgICAgICAgICAg ICAgICBleHBlY3RlZF9zdGRvdXQ9ZXhwZWN0ZWRfc3Rkb3V0LCBleHBlY3RlZF9zdGRlcnI9ZXhw ZWN0ZWRfc3RkZXJyKQorICAgICAgICBmaW5hbGx5OgorICAgICAgICAgICAgYnVpbGRlcnMuX2V4 YWN0X21hdGNoZXMgPSBvbGRfZXhhY3RfbWF0Y2hlcworCisgICAgZGVmIHRlc3RfcmViYXNlbGlu ZV90ZXN0X2ludGVybmFsX3dpdGhfbW92ZV9vdmVyd3JpdHRlbl9iYXNlbGluZXNfdG8oc2VsZik6 CisgICAgICAgIG9sZF9leGFjdF9tYXRjaGVzID0gYnVpbGRlcnMuX2V4YWN0X21hdGNoZXMKKyAg ICAgICAgdHJ5OgorICAgICAgICAgICAgYnVpbGRlcnMuX2V4YWN0X21hdGNoZXMgPSB7CisgICAg ICAgICAgICAgICAgIk1PQ0sgTGVvcGFyZCI6IHsicG9ydF9uYW1lIjogInRlc3QtbWFjLWxlb3Bh cmQiLCAic3BlY2lmaWVycyI6IHNldChbIm1vY2stc3BlY2lmaWVyIl0pfSwKKyAgICAgICAgICAg ICAgICAiTU9DSyBTbm93TGVvcGFyZCI6IHsicG9ydF9uYW1lIjogInRlc3QtbWFjLXNub3dsZW9w YXJkIiwgInNwZWNpZmllcnMiOiBzZXQoWyJtb2NrLXNwZWNpZmllciJdKX0sCisgICAgICAgICAg ICB9CisKKyAgICAgICAgICAgIGNvbW1hbmQgPSBSZWJhc2VsaW5lVGVzdCgpCisgICAgICAgICAg ICB0b29sID0gTW9ja1Rvb2woKQorICAgICAgICAgICAgdG9vbC5leGVjdXRpdmUgPSBNb2NrRXhl Y3V0aXZlKHNob3VsZF9sb2c9VHJ1ZSkKKyAgICAgICAgICAgIGNvbW1hbmQuYmluZF90b190b29s KHRvb2wpCisKKyAgICAgICAgICAgIHBvcnQgPSB0b29sLnBvcnRfZmFjdG9yeS5nZXQoJ3Rlc3Qt bWFjLXNub3dsZW9wYXJkJykKKyAgICAgICAgICAgIHRvb2wuZmlsZXN5c3RlbS53cml0ZV90ZXh0 X2ZpbGUodG9vbC5maWxlc3lzdGVtLmpvaW4ocG9ydC5iYXNlbGluZV92ZXJzaW9uX2RpcigpLCAn ZmFpbHVyZXMnLCAnZXhwZWN0ZWQnLCAnaW1hZ2UtZXhwZWN0ZWQudHh0JyksICcnKQorCisgICAg ICAgICAgICBvcHRpb25zID0gTW9ja09wdGlvbnMob3B0aW1pemU9VHJ1ZSwgYnVpbGRlcj0iTU9D SyBTbm93TGVvcGFyZCIsIHN1ZmZpeGVzPSJ0eHQiLAorICAgICAgICAgICAgICAgIG1vdmVfb3Zl cndyaXR0ZW5fYmFzZWxpbmVzX3RvPVsidGVzdC1tYWMtbGVvcGFyZCJdLCB2ZXJib3NlPVRydWUs IHRlc3Q9ImZhaWx1cmVzL2V4cGVjdGVkL2ltYWdlLmh0bWwiKQorCisgICAgICAgICAgICBvYyA9 IE91dHB1dENhcHR1cmUoKQorICAgICAgICAgICAgb2MuY2FwdHVyZV9vdXRwdXQoKQorICAgICAg ICAgICAgdHJ5OgorICAgICAgICAgICAgICAgIGxvZ3MgPSAnJworICAgICAgICAgICAgICAgIGNv bW1hbmQuZXhlY3V0ZShvcHRpb25zLCBbXSwgdG9vbCkKKyAgICAgICAgICAgIGZpbmFsbHk6Cisg ICAgICAgICAgICAgICAgXywgXywgbG9ncyA9IG9jLnJlc3RvcmVfb3V0cHV0KCkKKworICAgICAg ICAgICAgc2VsZi5hc3NlcnRUcnVlKCJDb3B5aW5nIGJhc2VsaW5lIGZyb20gL3Rlc3QuY2hlY2tv dXQvTGF5b3V0VGVzdHMvcGxhdGZvcm0vdGVzdC1tYWMtc25vd2xlb3BhcmQvZmFpbHVyZXMvZXhw ZWN0ZWQvaW1hZ2UtZXhwZWN0ZWQudHh0IHRvIC90ZXN0LmNoZWNrb3V0L0xheW91dFRlc3RzL3Bs YXRmb3JtL3Rlc3QtbWFjLWxlb3BhcmQvZmFpbHVyZXMvZXhwZWN0ZWQvaW1hZ2UtZXhwZWN0ZWQu dHh0LlxuIiBpbiBsb2dzKQorCisgICAgICAgIGZpbmFsbHk6CisgICAgICAgICAgICBidWlsZGVy cy5fZXhhY3RfbWF0Y2hlcyA9IG9sZF9leGFjdF9tYXRjaGVzCmRpZmYgLS1naXQgYS9Ub29scy9T Y3JpcHRzL3dlYmtpdHB5L3Rvb2wvc2VydmVycy9nYXJkZW5pbmdzZXJ2ZXJfdW5pdHRlc3QucHkg Yi9Ub29scy9TY3JpcHRzL3dlYmtpdHB5L3Rvb2wvc2VydmVycy9nYXJkZW5pbmdzZXJ2ZXJfdW5p dHRlc3QucHkKaW5kZXggMjEyMzdiNDhhODA4MzY1YzEzM2Q4NTAwZDQxNjkzNTUwZDIxMTA1MC4u YTJmM2ZhYmM4ZDc3YzBhMGRiNmRjYWU5ZjQ5NmFkN2M1MjRmYjlmNyAxMDA2NDQKLS0tIGEvVG9v bHMvU2NyaXB0cy93ZWJraXRweS90b29sL3NlcnZlcnMvZ2FyZGVuaW5nc2VydmVyX3VuaXR0ZXN0 LnB5CisrKyBiL1Rvb2xzL1NjcmlwdHMvd2Via2l0cHkvdG9vbC9zZXJ2ZXJzL2dhcmRlbmluZ3Nl cnZlcl91bml0dGVzdC5weQpAQCAtMTIxLDkgKzEyMSwzIEBAIGNsYXNzIEdhcmRlbmluZ1NlcnZl clRlc3QodW5pdHRlc3QuVGVzdENhc2UpOgogICAgICAgICBzZWxmLl9wb3N0X3RvX3BhdGgoIi9y ZWJhc2VsaW5lYWxsIiwgYm9keT0neyJ1c2VyLXNjcmlwdHMvYW5vdGhlci10ZXN0Lmh0bWwiOnsi TU9DSyBidWlsZGVyIjogWyJ0eHQiLCJwbmciXX19JywgZXhwZWN0ZWRfc3RkZXJyPWV4cGVjdGVk X3N0ZGVyciAlICgnTU9DSyBidWlsZGVyJywgJyJ0eHQiLCJwbmciJyksIGV4cGVjdGVkX3N0ZG91 dD1leHBlY3RlZF9zdGRvdXQsIHNlcnZlcj1zZXJ2ZXIpCiAKICAgICAgICAgc2VsZi5fcG9zdF90 b19wYXRoKCIvcmViYXNlbGluZWFsbCIsIGJvZHk9J3sidXNlci1zY3JpcHRzL2Fub3RoZXItdGVz dC5odG1sIjp7Ik1PQ0sgYnVpbGRlciAoRGVidWcpIjogWyJ0eHQiLCJwbmciXX19JywgZXhwZWN0 ZWRfc3RkZXJyPWV4cGVjdGVkX3N0ZGVyciAlICgnTU9DSyBidWlsZGVyIChEZWJ1ZyknLCAnInR4 dCIsInBuZyInKSwgZXhwZWN0ZWRfc3Rkb3V0PWV4cGVjdGVkX3N0ZG91dCkKLQotICAgIGRlZiB0 ZXN0X3JlYmFzZWxpbmVfbmV3X3BvcnQoc2VsZik6Ci0gICAgICAgIGJ1aWxkZXJzLl9leGFjdF9t YXRjaGVzID0geyJNT0NLIGJ1aWxkZXIiOiB7InBvcnRfbmFtZSI6ICJ0ZXN0LW1hYy1sZW9wYXJk IiwgInNwZWNpZmllcnMiOiBzZXQoWyJtb2NrLXNwZWNpZmllciJdKSwgIm1vdmVfb3ZlcndyaXR0 ZW5fYmFzZWxpbmVzX3RvIjogWyJtb2NrLXBvcnQtZmFsbGJhY2siLCAibW9jay1wb3J0LWZhbGxi YWNrMiJdfX0KLSAgICAgICAgZXhwZWN0ZWRfc3RkZXJyID0gJ01PQ0sgcnVuX2NvbW1hbmQ6IFtc J2VjaG9cJywgXCdyZWJhc2VsaW5lLWpzb25cJ10sIGN3ZD0vbW9jay1jaGVja291dCwgaW5wdXQ9 eyJ1c2VyLXNjcmlwdHMvYW5vdGhlci10ZXN0Lmh0bWwiOnsiTU9DSyBidWlsZGVyIjogWyJ0eHQi LCJwbmciXX19XG4nCi0gICAgICAgIGV4cGVjdGVkX3N0ZG91dCA9ICI9PSBCZWdpbiBSZXNwb25z ZSA9PVxuc3VjY2Vzc1xuPT0gRW5kIFJlc3BvbnNlID09XG4iCi0gICAgICAgIHNlbGYuX3Bvc3Rf dG9fcGF0aCgiL3JlYmFzZWxpbmVhbGwiLCBib2R5PSd7InVzZXItc2NyaXB0cy9hbm90aGVyLXRl c3QuaHRtbCI6eyJNT0NLIGJ1aWxkZXIiOiBbInR4dCIsInBuZyJdfX0nLCBleHBlY3RlZF9zdGRl cnI9ZXhwZWN0ZWRfc3RkZXJyLCBleHBlY3RlZF9zdGRvdXQ9ZXhwZWN0ZWRfc3Rkb3V0KQo=