Re: optimizing fork/exec in vendor source

2008-10-11 Thread Eric Blake
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Jack Andrews on 10/10/2008 10:21 PM: >> Exactly how do you propose to implement posix_spawnattr_setsigdefault >> without understanding cygwin internals, and given the fact that native >> Windows API is woefully lacking in sigset_t coding?

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Christopher Faylor
On Sat, Oct 11, 2008 at 12:50:03PM +1100, Jack Andrews wrote: >On Sat, Oct 11, 2008 at 1:26 AM, Christopher Faylor wrote: There is no reason to include my email address in the body of your message. That's what email headers are for. >>>this seems to be an easy problem in cygwin (at least, in co

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Jack Andrews
eric wrote: > Exactly how do you propose to implement posix_spawnattr_setsigdefault > without understanding cygwin internals, and given the fact that native > Windows API is woefully lacking in sigset_t coding? in the startup code for the new process, i'd call signal() (or similar) to set the defa

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Eric Blake
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Jack Andrews on 10/10/2008 7:50 PM: > i'm naive when it comes to cygwin internals, but the obvious > similarity of posix_spawn to CreateProcess is seductive. can you give > me an example of one of the difficulties of implementing posix_sp

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Jack Andrews
On Sat, Oct 11, 2008 at 1:26 AM, Christopher Faylor <[EMAIL PROTECTED]> wrote: >>this seems to be an easy problem in cygwin (at least, in comparison to >>fork). and would fix one of the biggest problems with cygwin (or at >>least cygwin bash). who has to hold the copyright? > > It is not an "easy

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Christopher Faylor
On Sat, Oct 11, 2008 at 12:26:32AM +1100, Jack Andrews wrote: >eric wrote: >>No, changing bash to use CreateProcess is not the answer. The real >>question is whether someone with copyright assignment is willing to >>write posix_spawn[p], and write it more efficiently than forking, > >this seems to

RE: optimizing fork/exec in vendor source

2008-10-10 Thread Dave Korn
Jack Andrews wrote on 10 October 2008 14:27: > eric wrote: >> No, changing bash to use CreateProcess is not the answer. The real >> question is whether someone with copyright assignment is willing to write >> posix_spawn[p], and write it more efficiently than forking, > > this seems to be an eas

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Eric Blake
Jack Andrews gmail.com> writes: > > eric wrote: > > No, changing bash to use CreateProcess is not the answer. The real > > question is whether someone with copyright assignment is willing to write > > posix_spawn[p], and write it more efficiently than forking, > > this seems to be an easy prob

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Jack Andrews
eric wrote: > No, changing bash to use CreateProcess is not the answer. The real > question is whether someone with copyright assignment is willing to write > posix_spawn[p], and write it more efficiently than forking, this seems to be an easy problem in cygwin (at least, in comparison to fork).

Re: optimizing fork/exec in vendor source

2008-10-10 Thread Eric Blake
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Jack Andrews on 10/9/2008 11:25 PM: > in the example, i guess all that is done is to fork and exec ls, > grabbing its stdout. this can be done with CreateProcess() easily > without having to emulate everything that fork does. so my real