Closed #357 via b8290ea5f671630a57c371950c342d58684451e8.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/357#event-1120591971
--
Reviewed by: George Wilson
Reviewed by: Matthew Ahrens
Reviewed by: Prashanth Sreenivasa
We need to initialize the dn_origin_obj_refd field of a dnode when it
is allocated from the kmem cache (or, alternatively, when it is
Closed #394 via 1d2bb1c86c216c433219772052d19bf0871cc985.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/394#event-1120368347
--
Reviewed by: Prakash Surya
Reviewed by: George Wilson
The problem is that zfs_get_data() supplies a stale zgd_bp to
dmu_sync(), which we then nopwrite against.
zfs_get_data() doesn't hold any DMU-related locks, so after it
copies db_blkptr
Reviewed by: Paul Dagnelie
Reviewed by: Pavel Zakharov
Reviewed by: George Wilson
The problem is that when dsl_bookmark_destroy_check() is
executed from open context (the pre-check), it fills in
dbda_success based on the
@arekinath Neat, I think that'd work well. The only downsides I see of using an
`NVList` directly are:
- Uniqueness checking is an `O(# of keys)` operation (whereas Lua tables use a
hash table). I'm sure we could build a workaround to deal with this.
- People reading Lua documentation to learn
In nvs_embedded(), when we return EINVAL due to reaching
the recursion limit, we should free the nvpriv_t that was
allocated earlier in the function.
Reviewed by: Pavel Zakharov
Reviewed by: George Wilson
Reviewed by: Prashanth Sreenivasa
Instead of having Lua return a table and converting that in C to an nvlist, why
not make the Lua code build the nvlist?
You can push in a C function (call it `NVList`) that returns a userdata
(pointing at a struct with an nvlist pointer in it) with a metatable that has
`__index` and
Closed #389 via b9cad68c52caf6f9f35813eb83c98ea4b96b17d0.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/389#event-1120052243
--
@dankimmel @ikozhukhov That's right, our take is that the core LUA code is
stable (changing only very slowly), so there would be little need to update it
in the future.
FYI, Here are the response that @cwill provided to the this question previously
@ikozhukhov: I believe some of your questions were already answered in the
previous thread (on https://github.com/openzfs/openzfs/pull/198). Regarding
updating the Lua interpreter, we will update it opportunistically if there's a
new feature or bug/security fix that we need, but otherwise there
about LUA - it is 3rd party library?
what version?
do you have your own fixes for it?
what are fixes - if they have?
(will be better to see special fixes over LUA as separate commit)
how you plans update it if needed?
--
You are receiving this because you are subscribed to this thread.
Reply to
grwilson approved this pull request.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/310#pullrequestreview-43429326
--
openzfs-developer
grwilson commented on this pull request.
> ARCSTAT_INCR(arcstat_l2_size, write_sz);
- ARCSTAT_INCR(arcstat_l2_asize, write_asize);
- vdev_space_update(dev->l2ad_vdev, write_asize, 0, 0);
+ ARCSTAT_INCR(arcstat_l2_asize, write_psize);
I agree. These variables and their
@avg-I pushed 1 commit.
25c6409 l2arc_write_buffers: rename size -> psize for better clarity
--
You are receiving this because you are subscribed to this thread.
View it on GitHub:
15 matches
Mail list logo