This is the mail archive of the libc-hacker@sourceware.cygnus.com mailing list for the glibc project.

Note that libc-hacker is a closed list. You may look at the archives of this list, but subscription and posting are not open.


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

Re: A patch for sunrpc


On Fri, Mar 31, 2000 at 08:35:40PM -0500, Roland McGrath wrote:
> > On Fri, Mar 31, 2000 at 08:24:15PM -0500, Roland McGrath wrote:
> > > It seems to me that if you are going to diddle with this then what you
> > > should do is make it more generally robust, i.e. grow the table.
> > > It already copes with using poll instead or in addition to select
> > > when the highest fd reaches FD_SETSIZE.  I'd also like to know why
> > 
> > I have thought about it. The main problem is struct, fd_set. How do you
> > change that at the runtime?
> 
> Well, it's already the case that fd_set can have variable size if
> everything supports it.  But anyway, you don't have to.  You can just use
> poll, and that is what the sunrpc code already does (it tests fd's against
> FD_SETSIZE).  It uses getdtablesize to size another table, which is a
> dubious prospect because in all current systems getdtablesize returns the
> soft resource limit (RLIMIT_NOFILE), which the program might increase with
> setrlimit after the first call into the rpcsvc stuff.


I forgot to mention that my patch is for glibc 2.1, which still uses
select for rpc, since it is what we are using today. We have no plan
to move glibc 2.2 any time soon. But we do want to support larger
NR_OPEN.


-- 
H.J. Lu (hjl@gnu.org)

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