Re: [Qemu-devel] [PATCH 3/3] net: EAGAIN handling for net/socket.c TCP

[Paolo Bonzini <pbonzini@redhat.com>]

Il 20/08/2012 16:57, Peter Maydell ha scritto:
> On 20 August 2012 14:27, Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> wrote:
>> > Replace spinning send_all() with a proper non-blocking send.  When the
>> > socket write buffer limit is reached, we should stop trying to send and
>> > wait for the socket to become writable again.
>> >
>> > +
>> > +    do {
>> > +        ret = writev(s->fd, iov, iovcnt);
>> > +    } while (ret == -1 && ret == EINTR);
> Quick sanity check -- does Windows have readv()/writev()? I notice
> that our slirp/slirp_config.h doesn't define HAVE_READV...

No, but we have

#define iov_recv(sockfd, iov, iov_cnt, offset, bytes) \
  iov_send_recv(sockfd, iov, iov_cnt, offset, bytes, false)
#define iov_send(sockfd, iov, iov_cnt, offset, bytes) \
  iov_send_recv(sockfd, iov, iov_cnt, offset, bytes, true)

that we can use instead.

Paolo