This is the mail archive of the libc-alpha@sources.redhat.com mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: openoffice 1.0.1 vs glibc cvs


Hi,

That should work for jdk 1.3.1 assuming those are the only missing symbols.  
I am rebuilding with that change as I write this.

I have no idea if similar changes will ever be made or released by 
Blackdown for JDK 1.1.8 and/or JDK 1.2.X for any platform.

One question: 

Why remove the functionality of being able to make glibc calls that avoid 
the pthread cancellation wrappers *before* the new pthreads you are 
designing is actually put into general use (the new pthreads would make 
the need to use those symbols to hack around non-posix signal handling in 
threads obselete).

Is there any way I could convince you to keep them *just until* your new 
pthreads is ready and being used on multiple platforms?  That way 
Blackdown can just revert to the the original solaris posix thread code 
that is part of the JDK and none of the Linux specific thread/signal hacks 
would be needed anymore.

The reason I ask is that jdk 1.2.X under x86 linux will have the same 
issues that jdk 1.3.1 under ppc linux has as well.

Thanks,

Kevin


 
On November 15, 2002 04:00, Ulrich Drepper wrote:
> Kevin Hendricks wrote:
> > 2.  somehow be able to call waitpid and wait from a signal handler
> > WITHOUT it being wrappered
> > by the pthread library and made a cancellation point.
>
> Make your own system calls.  Then you're responsible for it.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]