On 10/3/17 3:06 AM, Dave Chinner wrote:
> Hi Eric,
>
> I've put the latest mkfs refactor code that I have up in place
> you can pull it from. I've rebased it against the current for-next
> tree (4.13.1 release) and fixed all the problems that xfstests
> exposes. The only thing I haven't fixed is xfs/191 that does mkfs
> command line behaviour verification because the refactored version
> fixes several problems that the old mkfs didn't handle correctly
> (e.g. being able to specify certain things like agsize in blocks or
> sectors).
>
> There's a small filter patch needed for xfstests that I'll post in
> a reply to this pull request that will filter out the new "defaults
> sourced from ..." output and so prevent spurious xfstests failures.
>
> If you want I can tag the branch with a signed tag for you to pull
> from (same process as Linus prefers) rather than just a branch in a
> tree. If you'd prefer that I post this as patches instead, then let
> me know and I'll bomb the list instead.
Thanks Dave. Trying to un-swamp from other things.
Only advantage to patchbombing is having a permanent record of discussions,
but maybe it can happen in reference to patches as follows.... mostly
minor nits.
> mkfs: can't specify sector size of internal log
commit log is a bit vague/misleading; I think you mean that
it should be /disallowed/ because it must match the filesystem's
/sector/ size when it's internal? It's not actually clear to me why
an external log can have a "mismatched" sector size but internal can't.
> mkfs: introduce a structure to hold CLI options
"between th einput" (the input)
Comment mentions initialization to -1 or 0 but I don't see that this
gets done, at least not in this patch? (Am I missing it?) I took
this to mean that before anything happens this would be initialized
so we know what was and was not specified after everything was parsed...
> mkfs: introduce a structure to hold CLI options
+/* quick way of checking is a parameter was set on the CLI */
"if"
> mkfs: factor naming subopts parser
no "nvflag" in "temp don't break" code? Oh, nice - write-only var. Ok.
do we still test it for respec, or care? Will check later.
> mkfs: Introduce mkfs configuration structure
would be nice to proofread commit log a bit ;)
> mkfs: Introduce mkfs configuration structure
s/teh/the/
<ok stopping here for now, will send more later>
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
