Thanks!

This test is extremely flaky on Windows - it looks like it passes on the bot less than 10% of the time: http://lab.llvm.org:8011/buildslaves/win-py3-buildbot.

It looks like this broke the Windows bot. I am no longer receiving emails when the bot fails (as bot owner I used to always get the failure emails), so I am assuming for some reason notifications aren't working correctly.

LGTM

LGTM

I don't think that this would apply very well to Visual Studio with the current change because it is likely that if you are using VS to build LLDB, you also used VS to build the dependencies. Even for xcode, I am sure that there are people who build the dependencies standalone with xcode and then are building LLDB standalone with xcode, so any change that absolutely treats the scenario as one or the other is going to break someone.

I am not very familiar with the lldb-mi tests specifically, but it looks like they require python to run (and are marked as unsupported otherwise):

This change broke one of the tests on the Windows LLDB bot:

This caused failures on the windows bot again:

This looks good - it is still not all though. Even after Jonas' change the build failed with the same error elsewhere.

It looks like this was not enough and the build is still broken. Please revert your change or fix it and then keep an eye on the Buildbot to make sure it turns to green again. Once it's broken, it won't continue sending failure emails, so you have to monitor it to make sure your change fixed it.

Either this change or your other APValue change broke the windows bot:

Thanks, Pavel! I tested out the proposed change and all of our Ubuntu bots now work. I've committed it as well.

In D61611#1496838, @probinson wrote:

@stella.stamenova I'm not familiar with any lit feature that gives a special meaning to the prefix "no". The opposite of "REQUIRES: windows" is not "REQUIRES: nowindows" but "UNSUPPORTED: windows" AFAIK.
This part of the discussion should probably be taken to llvm-dev, though.

FTR I don't see that lldb's lit.cfg.py sets any features based on host OS, so even "UNSUPPORTED: windows" probably does not work currently.

In D61611#1496681, @sgraenitz wrote:

In D61611#1496644, @probinson wrote:

Sorry for the drive-by... what is this REQUIRES: nowindows? I don't see where lit generates this property. I grepped all of llvm-project.git and I see it used in two tests, but not where it's produced. Which suggests that those tests don't actually run *anywhere*.

Thanks for the heads-up. I also just grepped. The history says that Stella will know more about it. I went with XFAIL for now.

A couple of the tests from TestVSCode_attach.py (test_by_pid and test_by_name) are failing for us on Ubuntu because they are failing to attach: AssertionError: False is not True : attach failed (Operation not permitted). It looks like attaching by pid or by name requires elevation - if I rerun the same tests with sudo, they pass reliably. How did you run the tests when they passed for you?

Yes, sorry. I hadn't gotten around to testing it to make sure the failures were gone. Do you still need someone to commit for you?

I ran local tests with this and a couple of the tests from the lldb suite failed:

yay!

In D60230#1454293, @ruiu wrote:

Does this mean a commit to lld could potentially break lldb?

Thanks!

In D60122#1453462, @stella.stamenova wrote:

This broke the LLDB tests on the Windows Bot:

http://lab.llvm.org:8011/builders/lldb-x64-windows-ninja/builds/3237

This broke the LLDB tests on the Windows Bot:

LGTM

In D59847#1446671, @shafik wrote:

@stella.stamenova I committed a fix, please let me know if this does not address the regression:

http://llvm.org/viewvc/llvm-project?view=revision&revision=357210

This is causing failures on the windows bot. Please fix it or revert it.

This broke the Windows bot because now one of the TestPaths tests passes on Windows:

This caused a failure on the Windows bot:

This introduced a build break on Windows (which happened to coincide with another build break).

This is causing a failure on the Windows Bot:

This broke the windows LLDB bot:

This broke the windows bot:

LGTM. Now might not be the best time to commit it though because the LLDB build is broken on Windows.

In D58177#1397873, @labath wrote:

Thanks for the review Stella. I was hoping someone would step in and tell me that there is a better way to do that. However :), I see two problems with your proposal:

• build.py is a standalone file, and so it is not easy for it to share code with other stuff. This can be solved with some use_lldb_suite magic or moving the script some place else, but then comes the second problem:
• I am not sure these two use cases are actually compatible. Here I specifically want to "reinterpret_cast" binary bytes without doing any sort of encoding. The only place where to_string is used in build.py currently is on a value retrieved from the windows registry. I'm not 100% sure, but it seems to me that using utf8 encoding (which is what to_string is doing) is precisely the right thing to do there.
  
  With that in mind, I propose to do the following:
• leave build.py alone for now
• rename the new functions I'm introducing here to something less generic. Maybe bitcast_to_string and bitcast_to_bytes?
  
  WDYT?

This looks to have broken the windows build:

There's another patch (https://reviews.llvm.org/D54386/) which is fixing the fedora bot. You might want to leave that one.

This looks good. After it's checked in, it won't take effect until the master buildbot is updated though. You'd need @gkistanova to update it. She should also sign off on this.

This is failing on the windows bot:

This change broke one of the tests on Windows:

Build passes with VS now.

I am not sure if you saw this, but it looks like it broke one of the tests on the Windows bot:

@thakis Thanks for the information, I had actually missed that.

This change is causing build failures on Windows when using Visual Studio as the generator. It would also cause the same failure on for any other generator that supports multiple configurations.

I don't have a strong preference between using a separate module for interopability between python versions and implementing the functions as part of dotest if there are few of them, but I have to agree with @zturner that if we have more than a couple of functions, it makes sense to add them to an explicit module.

It looks like this change broke the windows bot:

In D57186#1372327, @craig.topper wrote:

@stella.stamenova I don't think reverting mine fixed it. I notice this commit was also around the same time, but isn't logged. http://llvm.org/viewvc/llvm-project?revision=352254&view=revision Does the bot also build lld and use it?

I haven't had time to look into it in details but it looks like either this change or https://reviews.llvm.org/D57259 broke the windows LLDB bot.

I haven't had time to look into it in details but it looks like either this change or https://reviews.llvm.org/D57186 broke the windows bot.