Re: [swift-dev] Flaky test - objc_nonnull_lie_hack

Oh, cool! The change looks good to me. It’s a bit of a hack but less so than -enable-source-import. There are annoying differences in behavior between -enable-source-import and the real module loader, and this is made worse by the fact that so many tests use -enable-source-import so we’re effec

Re: [swift-dev] Flaky test - objc_nonnull_lie_hack

Dmitri tracked it down and has a fix for it, and then I had a complaint about his fix, and then I didn't get back to it. :-) https://github.com/apple/swift/pull/4622 Jordan > On Sep 9, 2016, at 14:50, Slava Pestov via swift-dev > wrote: > > Running it in a loop on another machine doesn’t tr

Re: [swift-dev] CI outage

Master has been locked, we will send out another email once its open again. Thanks, Mishal Shah > On Sep 9, 2016, at 3:39 PM, Michael Gottesman via swift-dev > wrote: > > FYI, the tree will actually be locked until Sunday evening. This is so that > we can perform the necessary steps to move

Re: [swift-dev] CI outage

FYI, the tree will actually be locked until Sunday evening. This is so that we can perform the necessary steps to move swift to run on a newer LLVM/Clang, as announced last week. Michael > On Sep 9, 2016, at 10:49 AM, Nicole Jacque via swift-dev > wrote: > > Due to an HVAC maintenance outage

Re: [swift-dev] Flaky test - objc_nonnull_lie_hack

Running it in a loop on another machine doesn’t trigger it, so I’m thinking it might be a random variation in how the standard library is built? > On Sep 9, 2016, at 2:49 PM, Mark Lacey wrote: > > >> On Sep 9, 2016, at 2:45 PM, Slava Pestov via swift-dev > > wrote:

Re: [swift-dev] Flaky test - objc_nonnull_lie_hack

> On Sep 9, 2016, at 2:45 PM, Slava Pestov via swift-dev > wrote: > > Hi all, > > Is anyone else seeing this? I haven’t seen it in CI, but for me this test > fails at random, very rarely too — looks like something is non-deterministic > in the optimizer? Yes, I hit this a couple days ago an

Re: [swift-dev] Flaky test - objc_nonnull_lie_hack

I saw this once on Tuesday. I haven’t seen it since then. > On Sep 9, 2016, at 2:45 PM, Slava Pestov via swift-dev > wrote: > > Hi all, > > Is anyone else seeing this? I haven’t seen it in CI, but for me this test > fails at random, very rarely too — looks like something is non-deterministic

[swift-dev] Flaky test - objc_nonnull_lie_hack

Hi all, Is anyone else seeing this? I haven’t seen it in CI, but for me this test fails at random, very rarely too — looks like something is non-deterministic in the optimizer? /Users/slava/new/swift/test/SILGen/objc_nonnull_lie_hack.swift:30:9: error: expected string not found in input // OPT

Re: [swift-dev] Swift 2.3 branch build broken

+Nicole, Ted Can we delete swift-2.3-branch in swift-clang/swift-llvm? This is a common gotcha when checking out 2.3. If it’s not already tagged (which would be surprising), we can tag this for archival. Ben > On Sep 9, 2016, at 1:39 PM, Devin Coughlin via swift-dev > wrote: > > >> On Se

Re: [swift-dev] Swift 2.3 branch build broken

> On Sep 9, 2016, at 11:50 AM, Matt Rajca via swift-dev > wrote: > > Hello, > > I’ve been unable to build Swift 2.3 with both Xcode 8 GM on 10.12 and Xcode > 7.3/8 GM on 10.11. > > Trying to build with Xcode 8 GM on 10.12 and 7.3 GM on 10.11 results in: > "error: protocol '__BridgedNSError'

[swift-dev] Swift 2.3 branch build broken

Hello, I’ve been unable to build Swift 2.3 with both Xcode 8 GM on 10.12 and Xcode 7.3/8 GM on 10.11. Trying to build with Xcode 8 GM on 10.12 and 7.3 GM on 10.11 results in: "error: protocol '__BridgedNSError' requires '_NSErrorDomain' to be available on OS X 10.9.0 and newer” errors. Buildin

Re: [swift-dev] [Swift CI] Build Failure: 0. OSS - Swift Incremental RA - OS X (master) #6213

Hi John, Merged, thanks for fixing the issue. Mishal Shah > On Sep 9, 2016, at 11:16 AM, John McCall via swift-dev > wrote: > >> On Sep 9, 2016, at 10:51 AM, John McCall via swift-dev > > wrote: >> >>> On Sep 9, 2016, at 10:46 AM, no-re...@swift.org

Re: [swift-dev] [Swift CI] Build Failure: 0. OSS - Swift Incremental RA - OS X (master) #6213

> On Sep 9, 2016, at 10:51 AM, John McCall via swift-dev > wrote: > >> On Sep 9, 2016, at 10:46 AM, no-re...@swift.org >> wrote: >> [FAILURE] oss-swift-incremental-RA-osx [#6213] >> >> Build URL: https://ci.swift.org/job/oss-swift-incremental-RA-osx/6213/ >>

Re: [swift-dev] [Swift CI] Build Failure: 0. OSS - Swift Incremental RA - OS X (master) #6213

> On Sep 9, 2016, at 10:46 AM, no-re...@swift.org wrote: > [FAILURE] oss-swift-incremental-RA-osx [#6213] > > Build URL:https://ci.swift.org/job/oss-swift-incremental-RA-osx/6213/ > > Project: oss-swift-incremental-RA-osx > Da

[swift-dev] CI outage

Due to an HVAC maintenance outage in our lab, we will be taking down the CI systems starting a little after 4 PM California time, and turning them back on on Saturday morning. We will also be locking master during this time as pull requests will be unavailable. We’ll send out notification when

Re: [swift-dev] SIL to IR lowering

> On Sep 8, 2016, at 1:17 PM, Rebecca Chen via swift-dev > wrote: > > I was wondering where the SIL to IR lowering code is in the Swift source > code. That’s the responsibility of IRGen: https://github.com/apple/swift/tree/master/lib/IRGen > I am specifically looking for the code t

[swift-dev] SIL to IR lowering

I was wondering where the SIL to IR lowering code is in the Swift source code. I am specifically looking for the code that lowers the function return types. For example, from Int to i64. Any hints would be appreciated. ___ swift-dev mailing list swift

Re: [swift-dev] Cross Repository Testing with Swift CI

Mishal, this is absolutely fantastic. Excellent work, thank you! - Brian Gesiak On Thu, Sep 8, 2016 at 7:48 PM, mishal_shah via swift-dev < swift-dev@swift.org> wrote: > Cross Repository Testing with Swift CI > > Swift CI now supports the ability to test multiple pull request across > repositor