- User Since
- Aug 2 2014, 12:45 PM (138 w, 3 d)
Thu, Mar 23
Thanks. Generally speaking I like the change. Two things still bother me. First, as mav@ mentioned, ctld(8) will automatically load ctl(4) kernel module; it would be nice if things kept working for people depending on it. Second - that's a minor one, but the iSCSI frontend was introduced in 10.0, not 12.0.
The AUTHORS section for cfiscsi.4 seems completely wrong. Can you just copy it from iscsi.4?
Sun, Mar 19
Sat, Mar 18
Tue, Feb 28
Committed by wblock@ as r50019.
Reroot part moved to https://reviews.freebsd.org/D9834.
Note that parts of this functionality are not yet implemented.
Mon, Feb 27
Hm, something is wrong:
Sun, Feb 26
Feb 25 2017
It looks like it's still missing the update to usage() in ctladm.c, and the ctladm.8 manual page. The rest looks nice, thanks for working on this :-)
So, hm, is it good to go? Warren?
Feb 22 2017
Thanks. One thing I'm not sure about is the "USB OTG" stuff. "OTG" is not an acronym it itself, I think; "USB OTG" is. As in, the whole stuff.
Incorporate fixes from wblock@.
Feb 21 2017
Feb 20 2017
Bring back the correct diff.
Handle 32 bits as well. Note that this is completely untested.
Handle 32 bits as well. Note that the 32 bit part is untested.
Perhaps we could just add an additional check to not do this in case of softdep, ie if the buf has dependencies? Or rather, just add a conditional a bit lower to not call buf_deallocate() of them?
My worry here is that with this patch, we won't retry fsync if it failed with EIO, while with the previous code, due to the logic error, we would (at least if the failed buf was not the last one). I'm not sure if anything depends on it, though.
That's true, softdeps code has an explicit workaround for ENXIO. That said, I don't think we need to worry about softdep at all at this time - if it panics, it panics; it's not a regression, because it always panicked. It would be nice if UFS without softdep, or msdosfs, didn't panic, and for that, we need to handle those bufs somewhere.
Feb 19 2017
Feb 18 2017
Remove the pointless wrapper.
The panic was changed into printf already, as it triggered every time, even when UFS or msdosfs could survive just fine. The warning - I kind of agree, it could be useful, but I don't think it's worth complicating the code just for that.
Feb 17 2017
Konstantin - _all_ the utilities. Namely newfs and mount. They all default to ufs. It's logical for fsck to do the same. Especially that other systems people are used to don't require one to specify the filesystem by hand either.
Feb 16 2017
Wouldn't that defeat the purpose of having the default? Other utilities don't issue warnings in this case, they just silently go with default.