On 02/11/15 14:24, Ian Campbell wrote:
> On Mon, 2015-11-02 at 14:10 +, Andrew Cooper wrote:
>> On 02/11/15 13:44, Ian Campbell wrote:
>>> On Tue, 2015-10-27 at 17:10 +, Wei Liu wrote:
When oxenstored wrote to the ring, it wrote a chunk of contiguous
data.
Originally when it
On Mon, 2015-11-02 at 14:10 +, Andrew Cooper wrote:
> On 02/11/15 13:44, Ian Campbell wrote:
> > On Tue, 2015-10-27 at 17:10 +, Wei Liu wrote:
> > > When oxenstored wrote to the ring, it wrote a chunk of contiguous
> > > data.
> > > Originally when it tried to write across ring boundary, it
On 02/11/15 13:44, Ian Campbell wrote:
> On Tue, 2015-10-27 at 17:10 +, Wei Liu wrote:
>> When oxenstored wrote to the ring, it wrote a chunk of contiguous data.
>> Originally when it tried to write across ring boundary, it returned a
>> short-write when there is still room. That led to stalli
On Tue, 2015-10-27 at 17:10 +, Wei Liu wrote:
> When oxenstored wrote to the ring, it wrote a chunk of contiguous data.
> Originally when it tried to write across ring boundary, it returned a
> short-write when there is still room. That led to stalling mini-os's
> xenstore thread at times.
Wh
On 27/10/15 17:10, Wei Liu wrote:
> When oxenstored wrote to the ring, it wrote a chunk of contiguous data.
> Originally when it tried to write across ring boundary, it returned a
> short-write when there is still room. That led to stalling mini-os's
> xenstore thread at times.
>
> Fix this by ca
On 28/10/15 13:34, David Scott wrote:
>
>> On 27 Oct 2015, at 17:31, Andrew Cooper wrote:
>>
>> On 27/10/15 17:28, Samuel Thibault wrote:
>>> Andrew Cooper, le Tue 27 Oct 2015 17:21:39 +, a écrit :
as the second attempted write could return short as well.
>>> That is fine. The second atte
David Scott, le Wed 28 Oct 2015 13:34:10 +, a écrit :
> However I believe the ‘suspicious’ OCaml patch fixes the specific issue (— or
> have I missed something?).
It does.
> Does anyone have time to prototype what a C-level fix would look like?
Here is a try, untested though.
Signed-off-by
> On 27 Oct 2015, at 17:31, Andrew Cooper wrote:
>
> On 27/10/15 17:28, Samuel Thibault wrote:
>> Andrew Cooper, le Tue 27 Oct 2015 17:21:39 +, a écrit :
>>> as the second attempted write could return short as well.
>> That is fine. The second attempt will only return a short write if there
Another way would be to make ml_interface_write write both pieces
instead of just a contiguous one. The caml code would then look less
suspicious :)
Samuel
___
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
On 27/10/15 17:28, Samuel Thibault wrote:
> Andrew Cooper, le Tue 27 Oct 2015 17:21:39 +, a écrit :
>> as the second attempted write could return short as well.
> That is fine. The second attempt will only return a short write if there
> was really not enough room for it, which is what we want.
Andrew Cooper, le Tue 27 Oct 2015 17:21:39 +, a écrit :
> as the second attempted write could return short as well.
That is fine. The second attempt will only return a short write if there
was really not enough room for it, which is what we want.
Samuel
__
Andrew Cooper, le Tue 27 Oct 2015 17:21:39 +, a écrit :
> The correct behaviour would be for Xs_ring.write to return the actual
> number of bytes it put into the ring, even if this is shorter than len.
That is already what it does.
Samuel
___
Xen-d
That being said,
Wei Liu, le Tue 27 Oct 2015 17:10:09 +, a écrit :
> @@ -91,10 +91,12 @@ let write_fd back con s len =
> Unix.write back.fd s 0 len
>
> let write_mmap back con s len =
> - let ws = Xs_ring.write back.mmap s len in
> - if ws > 0 then
> + let ws = ref (Xs_rin
On 27/10/15 17:10, Wei Liu wrote:
> When oxenstored wrote to the ring, it wrote a chunk of contiguous data.
> Originally when it tried to write across ring boundary, it returned a
> short-write when there is still room. That led to stalling mini-os's
> xenstore thread at times.
>
> Fix this by cal
When oxenstored wrote to the ring, it wrote a chunk of contiguous data.
Originally when it tried to write across ring boundary, it returned a
short-write when there is still room. That led to stalling mini-os's
xenstore thread at times.
Fix this by calling write function for a second time when th
15 matches
Mail list logo