On 14.02.20 13:17, Eugenio Pérez wrote:

Can you try the inlined patch over 52c36ce7f334 ("vhost: use batched version by default")? My intention is to check if
"strange VHOST_SET_VRING_BASE" line appears. In previous tests, it appears very fast, but maybe it takes some time for
it to appear, or it does not appear anymore.

  LD [M]  drivers/vhost/vhost_vsock.o
  CC [M]  drivers/vhost/vhost.o
In file included from ./include/linux/printk.h:331,
                 from ./include/linux/kernel.h:15,
                 from ./include/linux/list.h:9,
                 from ./include/linux/wait.h:7,
                 from ./include/linux/eventfd.h:13,
                 from drivers/vhost/vhost.c:13:
drivers/vhost/vhost.c: In function ‘fetch_descs’:
drivers/vhost/vhost.c:2330:56: error: ‘struct vhost_virtqueue’ has no member named ‘first_desc’
 2330 |   vq, vq->last_avail_idx, vq->avail_idx, vq->ndescs, vq->first_desc);
      |                                                        ^~
./include/linux/dynamic_debug.h:125:15: note: in definition of macro ‘__dynamic_func_call’
  125 |   func(&id, ##__VA_ARGS__);  \
      |               ^~~~~~~~~~~
./include/linux/dynamic_debug.h:153:2: note: in expansion of macro ‘_dynamic_func_call’
  153 |  _dynamic_func_call(fmt, __dynamic_pr_debug,  \
      |  ^~~~~~~~~~~~~~~~~~
./include/linux/printk.h:335:2: note: in expansion of macro ‘dynamic_pr_debug’
  335 |  dynamic_pr_debug(fmt, ##__VA_ARGS__)
      |  ^~~~~~~~~~~~~~~~
drivers/vhost/vhost.c:2328:2: note: in expansion of macro ‘pr_debug’
 2328 |  pr_debug(
      |  ^~~~~~~~
make[2]: *** [scripts/Makefile.build:266: drivers/vhost/vhost.o] Error 1
make[1]: *** [scripts/Makefile.build:503: drivers/vhost] Error 2