[swift-dev] Major SwiftPM compile time regression

Mark Lacey mark.lacey at apple.com
Sun Sep 25 01:17:02 CDT 2016


I just pulled this up under Instruments and found that 94% of the compile time for StringMangling.swift.o, which seems to be where all the time goes, is spent in performLLVM, so this doesn’t look like a type checker issue.

Weight	Self Weight		Symbol Name
10.32 min   94.4%	1.00 ms	 	       performLLVM(swift::IRGenOptions&, swift::DiagnosticEngine&, llvm::sys::SmartMutex<false>*, llvm::GlobalVariable*, llvm::Module*, llvm::TargetMachine*, swift::version::Version const&, llvm::StringRef)

I attached a screen shot from Instruments to the bug.

Mark

> On Sep 24, 2016, at 9:25 PM, Daniel Dunbar <daniel_dunbar at apple.com> wrote:
> 
> I'm assuming its the type checker based on how it manifests (and we haven't changed anything interesting in the module that is slow to build), but I haven't looked at any samples.
> 
>  - Daniel
> 
>> On Sep 24, 2016, at 9:23 PM, Ted Kremenek <kremenek at apple.com <mailto:kremenek at apple.com>> wrote:
>> 
>> Does this look related to the type checker (I saw Mark’s comments in the SR) or something else?
>> 
>>> On Sep 24, 2016, at 3:23 PM, Daniel Dunbar via swift-dev <swift-dev at swift.org <mailto:swift-dev at swift.org>> wrote:
>>> 
>>> Swift TOT is currently taking a very long time (and upwards of 8GB) to build SwiftPM. I filed:
>>>  https://bugs.swift.org/browse/SR-2754 <https://bugs.swift.org/browse/SR-2754>
>>> can someone on the compiler take a look? This makes it hard to develop with TOT.
>>> 
>>> - Daniel
>>> 
>>> _______________________________________________
>>> swift-dev mailing list
>>> swift-dev at swift.org <mailto:swift-dev at swift.org>
>>> https://lists.swift.org/mailman/listinfo/swift-dev <https://lists.swift.org/mailman/listinfo/swift-dev>
>> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.swift.org/pipermail/swift-dev/attachments/20160924/b362dc41/attachment.html>


More information about the swift-dev mailing list