Re: writev() "bug"

Matthias Urlichs (smurf@smurf.noris.de)
Wed, 5 Jun 1996 20:17:59 +0100

Messages sorted by: [ date ][ thread ][ subject ][ author ]
Next message: Albert Cahalan: "Re: SVGA kernel chipset drivers."
Previous message: Jason McMullan: "Re: SVGA kernel chipset drivers."
Maybe in reply to: Marc Unangst: "writev() "bug""

In linux.dev.kernel, article <01I5GZVSVE2A000J4E@VAX1.ROCKHURST.EDU>,
Aaron Ucko <UCKO@vax1.rockhurst.edu> writes:
> >
> >Then libc is broke writev/readv makes atomicity guarantees.
>
> Well, it was even worse before the syscall existed, since then it could
> do only 1 iov at a time.
>
I don't think so. People have used INN for ages, and it makes extensive use
of writev.

-- Prosperity is not without many fears and distaste; adversity not without many comforts and hopes. -- Francis Bacon

-- 
Matthias Urlichs

Next message: Albert Cahalan: "Re: SVGA kernel chipset drivers."
Previous message: Jason McMullan: "Re: SVGA kernel chipset drivers."
Maybe in reply to: Marc Unangst: "writev() "bug""