882779 - Do not attempt to run the LLVM PR8927 test when building with Clang on Windows

Status: RESOLVED FIXED

(Firefox Build System :: General, defect)

Description

[(no longer active)]

Attachment: Patch (v1)

Comment 1

[(not currently active) Ted Mielczarek]

Comment on attachment 762136 [details] [diff] [review]
Patch (v1)

Review of attachment 762136 [details] [diff] [review]:
-----------------------------------------------------------------

::: configure.in
@@ +2886,4 @@
>  
>  MOZ_GCC_PR49911
>  MOZ_GCC_PR39608
> +if test -z "$MOZ_WINSDK_MAXVER"; then

This feels like a weird thing to test. Why not just test that OS_TARGET = "WINNT"?

Comment 2

[(no longer active)]

(In reply to Ted Mielczarek [:ted.mielczarek] from comment #1)
> Comment on attachment 762136 [details] [diff] [review]
> Patch (v1)
> 
> Review of attachment 762136 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> ::: configure.in
> @@ +2886,4 @@
> >  
> >  MOZ_GCC_PR49911
> >  MOZ_GCC_PR39608
> > +if test -z "$MOZ_WINSDK_MAXVER"; then
> 
> This feels like a weird thing to test. Why not just test that OS_TARGET =
> "WINNT"?

Because I pretty much don't know what I'm doing :D

Comment 3

[(no longer active)]

Attachment: Patch (v2)

Comment 4

[(not currently active) Ted Mielczarek]

Comment on attachment 765010 [details] [diff] [review]
Patch (v2)

Review of attachment 765010 [details] [diff] [review]:
-----------------------------------------------------------------

I will r+ this even though the patch is wrong because I trust you to fix it.

::: configure.in
@@ +2886,4 @@
>  
>  MOZ_GCC_PR49911
>  MOZ_GCC_PR39608
> +if test "$OS_TARGET" = WINNT; then

Uh, you clearly want to test != here. (Are you testing these patches?)

Comment 5

[(no longer active)]

(In reply to Ted Mielczarek [:ted.mielczarek] from comment #4)
> Comment on attachment 765010 [details] [diff] [review]
> Patch (v2)
> 
> Review of attachment 765010 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> I will r+ this even though the patch is wrong because I trust you to fix it.
> 
> ::: configure.in
> @@ +2886,4 @@
> >  
> >  MOZ_GCC_PR49911
> >  MOZ_GCC_PR39608
> > +if test "$OS_TARGET" = WINNT; then
> 
> Uh, you clearly want to test != here. (Are you testing these patches?)

Sorry, I first typed in the change on my Windows box, and then I edited the patch on my Mac to attach it to the bug.  :-)  It did work with clang on Windows so my guess is that I messed up something while I was doing that...

https://hg.mozilla.org/integration/mozilla-inbound/rev/3cc992814a7e

Comment 6

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/3cc992814a7e

Comment 7

[(no longer active)]

I reverted this because the hack is no longer needed:

https://hg.mozilla.org/integration/mozilla-inbound/rev/78930b6efe65