* Maxim Sobolev, 2016-02-19 :
> Thanks, Thomas. IMHO this is quite serious bug in one of the core
> utilities, so I suggest it gets faster MFC before 10.3 goes into beta stage
> on Feb 26. The old code in RELENG_10 is just plain broken anyway now, so it
> could not make it any worse. Thanks!
It t
Thanks, Thomas. IMHO this is quite serious bug in one of the core
utilities, so I suggest it gets faster MFC before 10.3 goes into beta stage
on Feb 26. The old code in RELENG_10 is just plain broken anyway now, so it
could not make it any worse. Thanks!
On Thu, Feb 18, 2016 at 12:44 AM, Thomas Qu
* Maxim Sobolev, 2016-02-10 :
> In any case, I've opened a bug here:
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=207092
Thanks Maxim, much appreciated. Candidate fix is on review:
https://reviews.freebsd.org/D5248
Thomas.
pgp3X8i_x0IZn.pgp
Description: PGP signature
The patch looks fair to me, although I have not got chance to test it yet.
You probably don't need += since cnt is already tested to be 0, so "="
would suffice. :)
In any case, I've opened a bug here:
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=207092
So please make sure to deal with it wh
* Maxim Sobolev, 2016-02-10 :
> Looking at the code in question I don't see how could it have worked. Look
> at the following piece of code in your diff for example:
>
> + if (force && cnt == 0) {
> + pending -= l
* Maxim Sobolev, 2016-02-10 :
> $ dd if=/dev/zero of=/tmp/foo.bar bs=1m count=10
> 10+0 records in
> 10+0 records out
> 10485760 bytes transferred in 0.003244 secs (3232431656 bytes/sec)
> $ ktrace dd if=/tmp/foo.bar of=/tmp/foo.bar1 bs=1m conv=sparse
> 10+0 records in
> 10+0 records out
> $ ls -l
Something is seriously broken about this change. The last block is never
written at least for the case of output being regular file:
$ dd if=/dev/zero of=/tmp/foo.bar bs=1m count=10
10+0 records in
10+0 records out
10485760 bytes transferred in 0.003244 secs (3232431656 bytes/sec)
$ ktrace dd if=/
