| Message ID | 1470241360-3574-4-git-send-email-berrange@redhat.com |
|---|---|
| State | New |
| Headers | show |
On Wed, 3 Aug 2016 17:22:38 +0100 "Daniel P. Berrange" <berrange@redhat.com> wrote: s/sclpconsole/sclpconsolelm/ in the subject, as there are two sclp consoles :) > The write_console_data() method in sclpconsole-lm.c checks > whether the return value of qemu_chr_fe_write() has the > value of -EAGAIN and if so then increments the buffer offset > by the value of EAGAIN. Fortunately qemu_chr_fe_write() will > never return EAGAIN directly, rather it returns -1 with > errno set to EAGAIN, so this broken code path was not > reachable. The behaviour on EAGAIN was stil bad though, > causing the write_console_data() to busy_wait repeatedly > calling qemu_chr_fe_write() with no sleep between iters. > > Just remove all this loop logic and replace with a call > to qemu_chr_fe_write_all(). > > Signed-off-by: Daniel P. Berrange <berrange@redhat.com> > --- > hw/char/sclpconsole-lm.c | 18 +++--------------- > 1 file changed, 3 insertions(+), 15 deletions(-) > > diff --git a/hw/char/sclpconsole-lm.c b/hw/char/sclpconsole-lm.c > index a22ad8d..8cb0026 100644 > --- a/hw/char/sclpconsole-lm.c > +++ b/hw/char/sclpconsole-lm.c > @@ -201,21 +201,9 @@ static int write_console_data(SCLPEvent *event, const uint8_t *buf, int len) > return len; > } > > - buf_offset = buf; > - while (len > 0) { > - ret = qemu_chr_fe_write(scon->chr, buf, len); > - if (ret == 0) { > - /* a pty doesn't seem to be connected - no error */ > - len = 0; > - } else if (ret == -EAGAIN || (ret > 0 && ret < len)) { > - len -= ret; > - buf_offset += ret; > - } else { > - len = 0; > - } > - } > - > - return ret; > + /* XXX this blocks entire thread. Rewrite to use > + * qemu_chr_fe_write and background I/O callbacks */ > + return qemu_chr_fe_write_all(scon->chr, buf, len); This is basically the same change that had been done for the sclp vt220 console already in 2e14211 ("s390/sclpconsole: handle char layer busy conditions"), so this looks fine. > } > > static int process_mdb(SCLPEvent *event, MDBO *mdbo) Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
diff --git a/hw/char/sclpconsole-lm.c b/hw/char/sclpconsole-lm.c index a22ad8d..8cb0026 100644 --- a/hw/char/sclpconsole-lm.c +++ b/hw/char/sclpconsole-lm.c @@ -201,21 +201,9 @@ static int write_console_data(SCLPEvent *event, const uint8_t *buf, int len) return len; } - buf_offset = buf; - while (len > 0) { - ret = qemu_chr_fe_write(scon->chr, buf, len); - if (ret == 0) { - /* a pty doesn't seem to be connected - no error */ - len = 0; - } else if (ret == -EAGAIN || (ret > 0 && ret < len)) { - len -= ret; - buf_offset += ret; - } else { - len = 0; - } - } - - return ret; + /* XXX this blocks entire thread. Rewrite to use + * qemu_chr_fe_write and background I/O callbacks */ + return qemu_chr_fe_write_all(scon->chr, buf, len); } static int process_mdb(SCLPEvent *event, MDBO *mdbo)
The write_console_data() method in sclpconsole-lm.c checks whether the return value of qemu_chr_fe_write() has the value of -EAGAIN and if so then increments the buffer offset by the value of EAGAIN. Fortunately qemu_chr_fe_write() will never return EAGAIN directly, rather it returns -1 with errno set to EAGAIN, so this broken code path was not reachable. The behaviour on EAGAIN was stil bad though, causing the write_console_data() to busy_wait repeatedly calling qemu_chr_fe_write() with no sleep between iters. Just remove all this loop logic and replace with a call to qemu_chr_fe_write_all(). Signed-off-by: Daniel P. Berrange <berrange@redhat.com> --- hw/char/sclpconsole-lm.c | 18 +++--------------- 1 file changed, 3 insertions(+), 15 deletions(-)