include ext4

Jason Newton nevion at
Tue Sep 23 20:21:28 UTC 2008

Scott James Remnant wrote:
> On Sat, 2008-09-20 at 04:41 -0700, Jason Newton wrote:
>> I'd also like to put my vote in on it being turned on
> I'd argue that since it's a development filesystem, people using it need
> to be ready to supply patches to the author for the problems they
> encounter.
> Thus forcing them to build their own kernel with it is entirely
> appropriate.
> Scott
While it may be true someone may need to submit patches, I think this is 
fairly unlikely, take a look at the ext4 bugzilla to see how many bugs 
have existed (of course these are only reported bugs and its true that a 
limited number of people currently test/use ext4):
Expanding the search criteria only shows a few more, almost all are 
fixed.  Its actually very similar to ext3's page.

As for the dev. mailing list - there is a flurry of patches, but almost 
all of the patches are for noncritical/featurish things, as such there 
just doesn't seem to be any show stoppers left and right like some might 
think.  Most importantly, the disk format is NOT changing anymore and 
just having ro support is a significant plus, if there is limited write 
support all the better, though it seems quite stable.  Finally, remember 
that this is a journaled filesystem, and like ext3, it supports 
different modes of journaling modes 
(data=journal,data=ordered,data=writeback)  meaning that even if there 
are hiccups (I've not experienced any yet, YMMV), its  pretty likely 
your data is still safe.

More information about the kernel-team mailing list