On Wed, Sep 16, 2015 at 8:25 AM, Jack Howarth <howarth.at.f...@gmail.com>
wrote:

>
>
> On Wed, Sep 16, 2015 at 5:43 AM, Max Horn <m...@quendi.de> wrote:
>
>> Jack,
>>
>> On 15.09.2015, at 03:18, Jack Howarth <howarth.at.f...@gmail.com> wrote:
>>
>> > Max,
>> >
>> > El Capitan GM currently produces an instability in fink make for a
>> variety of builds (cmake, libcurl4, texlive-base, r-base32, etc) resulting
>> in random failures of the form...
>> >
>> > make: INTERNAL: Exiting with 1 jobserver tokens available; should be 8!
>> >
>> > during parallel builds on my MacPro 3,1 with 8 cores.
>>
>> In a previous email you mentioned that MacPorts' make doesn't show this
>> problem? Does it apply any patches to make? And did you verify that your
>> test builds with MacPorts really used their GNU make, and not the system
>> make after all?
>>
>>
>>
> MacPorts gmake doesn't show the problem on the command line or in their
> builds but I have only been able to test a rather narrow set of MacPorts
> builds against their gmake. However I have been able to reproduce the build
> failures under fink when the call to 'make' is substituted with
> '/opt/local/bin/make' to invoke their copy.
>
> > Since only a few packages in fink have a dependency on fink make, those
>> could be updated to invoke it as gmake and allow the average user to keep
>> using the system make unless they really want to change the default using
>> make-default.
>>
>> I am not 100% comfortable with changing the behavior of the make package
>> like this -- people who are relying on "GNU make" will be caught by
>> surprise if the "make" package suddenly does not provide gmake under the
>> name "make" anymore. Also, packages which depend on "make" would have to be
>> changed at the exact same time as this change is applied (for this I'd need
>> to know which packages are affected, though, which I don't right now... do
>> we have an effective way to do that?)
>>
>> More workable would be to changes this just for 10.11, i.e. split make
>> into a pre-10.11 version and a 10.11+ version.
>>
>> Since I don't have 10.11, I can't do test that myself, and I am not even
>> sure whether 10.11 is going to use the 10.7 tree, or a fresh tree ? If it
>> does use 10.7, I could add a make-10.11.info based on what you sent,
>> with a revision of 1001 and "Distribution: 10.11"
>>
>>
>>
> I doubt a unique set of packages exists to protect from this bug with the
> usage of /usr/bin/make. The bug looks to be a race condition in the
> threading support as tweaking the optimization of make build modulates the
> number of packages that fail under fink make. Likewise reducing the number
> of cores in the parallel build can suppress the bug. Also, when I debugged
> this by adding -d, only one of the failing tests, texlive-base, remained
> and even then only randomly. So the real question will be how many users
> machines have the cores/speed to tickle the bug. My past experience with
> Apple and threading bugs is that it is futile to expect them to be fixed in
> a given OS release and usually you have to wait for the next major kernel
> rewrite.
>
>
>
One other comment. It isn't particularly surprising that this bug isn't
suppressed by disabling SIP...

https://developer.apple.com/library/prerelease/ios/documentation/Security/Conceptual/System_Integrity_Protection_Guide/System_Integrity_Protection_Guide.pdf

The previous kernel code path, which didn't have the ability to do kernel
level pruning of environmental variables from processes, is unlikely to be
restored simply by disabling SIP. More likely the just security pruning
itself is disabled.


> Cheers,
>> Max
>
>
>
------------------------------------------------------------------------------
Monitor Your Dynamic Infrastructure at Any Scale With Datadog!
Get real-time metrics from all of your servers, apps and tools
in one place.
SourceForge users - Click here to start your Free Trial of Datadog now!
http://pubads.g.doubleclick.net/gampad/clk?id=241902991&iu=/4140
_______________________________________________
Fink-devel mailing list
Fink-devel@lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Reply via email to