Feature #411
closedRemove BKL reliance in encode_caps_cb
0%
Description
This looks like it will require extending our pagelist API.
Updated by Greg Farnum over 13 years ago
- Status changed from New to 4
It hasn't been extensively tested but it doesn't break normal reconnect, at least. Waiting for Sage to take a look at it before packaging it up for upstream.
Updated by Sage Weil over 13 years ago
- Assignee changed from Greg Farnum to Sage Weil
Now in unstable branch.
Need to rebase all of unstable, push it into linux-next, and then send this to Arnd.
Updated by Sage Weil over 13 years ago
Actually I should check if git merge is smart enough to handle the pagelist rename queued for linux-next. If so, we can send Arnd the patches based on current upstream and not worry about the merge order. If not, it's annoying..
Updated by Sage Weil over 13 years ago
Greg, can you also write a quick test for the pagelist stuff? Maybe something along the lines of
x = rand() % 100000; append to pagelist for (i=0 to 1000) { set cursor y = rand() % 100000; append y bytes truncate }
I'm worried just because I fixed a few bugs looking at it yesterday with the kmap/kunmap, and the code path will be so so rarely executed any errors won't come up in our normal testing.
Updated by Greg Farnum over 13 years ago
- Assignee changed from Sage Weil to Greg Farnum