striptease build failures

NEW
Unassigned

Status

Camino Graveyard
General
6 years ago
6 years ago

People

(Reporter: philippe (part-time), Unassigned)

Tracking

1.9.2 Branch
x86
Mac OS X

Details

Attachments

(2 attachments)

(Reporter)

Description

6 years ago
Created attachment 588575 [details]
buildlog on 10.7 / gcc4.0 / 10.4u sdk

* breaks the build on 10.7 Lion with GCC 4.0 / 10.4u SDK / Xcode 4
* fails with the 10.5 SDK and GCC 4.2, at least on 10.6 / Xcode 3.2.x (I'll need to re-generate a build log for that)
(In reply to philippe from comment #0)

> * fails with the 10.5 SDK and GCC 4.2, at least on 10.6 / Xcode 3.2.x 

philippe reminded me we'd originally filed bug 569779 for that; per the comments there, it seems like it was the gcc-4.2 half of the equation that was the problem, because gcc-4.0.1 + 10.5 SDK apparently worked.
(In reply to comment #1)
> comments there, it seems like it was the gcc-4.2 half of the equation that
> was the problem, because gcc-4.0.1 + 10.5 SDK apparently worked.

Er, and if my analysis in bug 717513 comment 8, striptease isn't using the SDK at all (although mento implied otherwise in bug 414397 comment 1), so it definitely would have been just the compiler version in the gcc-4.2 + 10.5 SDK situation.
Bug 717513 comment 8 (regarding the 10.7 / gcc-4.0 / 10.4u SDK failures):

> Ah; the first failure, at least, is a libstuff file looking for a system
> header that's gone on 10.7 (see bug 673789 for a similar problem with
> Breakpad), and apparently the SDK selection isn't getting passed into the
> striptease build.  That's, I think, because striptease is built as host
> program; I'd have to root around in the flag setup to be sure.
> 
> The other errors may follow from that; it's not clear at a glance, and I
> don't have time to dig around there right now.
(Reporter)

Comment 4

6 years ago
Created attachment 588799 [details]
buildlog on 10.6 / gcc 4.2 /10/5 sdk
You need to log in before you can comment on or make changes to this bug.