This is the mail archive of the cygwin mailing list for the Cygwin 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: Use of Dual Core causes random failures building OpenJDK

On Nov 20 09:32, Christopher Faylor wrote:
> On Fri, Nov 20, 2009 at 12:20:52PM +0100, Corinna Vinschen wrote:
> >On Nov 20 06:04, mike marchywka wrote:
> >> On 11/19/09, Larry Hall (Cygwin) <> wrote:
> >>>Wow! I had a hunch that BLODA could be the cause of the file related
> >>>problems but it never occurred to me that it could be screwing up
> >>>environment variables too.
> >>
> >>Anyone care to determine if env updates are not thread safe?
> >
> >It's indeed not thread safe.  Looks like we need some locking...
> What's "it" in this case?  putenv?

"it" is primarily _addenv, which is used by setenv and putenv.

>   The putenv() function is not required to be reentrant, and the one in
>   libc4, libc5 and glibc 2.0 is not, but the glibc 2.1 version is

Same for setenv and unsetenv, which don't need to be thread-safe per
POSIX.  glibc's setenv is thread-safe, though.

> I agree that it would be nice to make it safer but is there really code
> here which is updating the environment in separate threads?

And given POSIX, if so, it would be an application bug if the application
doesn't care by itself to use setenv/getenv in a thread-safe manner.


Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

Problem reports:
Unsubscribe info:

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