ecryptfs hangs on kernel 2.6.22-7.14

Binary package hint: linux-source-2.6.22

On Thu, Jul 05, 2007 at 09:27:20AM -0300, William Lima wrote:
> The ecryptfs is hanging on kernel 2.6.22-rc7. Please check this link:
>
> https://bugs.launchpad.net/ubuntu/+source/ecryptfs-utils/+bug/122852

This is why eCryptfs is still marked ``EXPERIMENTAL'' in the kernel...

This is probably related to the series of llseek() patches that I've
been working on over the last month. To get eCryptfs back to the way
it was, the last 4 GIT commits need to be dropped:

53a2731f9310a66beaf55677229ab067c85ce4fa
240e2df5c740d73fc08cac9989872212deb2d20e
e10f281bca03f45bdec91e67645c394eaec2f8f6
d4c5cdb3e099da7cc64df622b02ff7659babe16e

Without these patches, eCryptfs will fill in 0's on an llseek()
without a write, which technically isn't correct POSIX behavior, but
it's better than hanging.

In the meantime, I'll add the mount-touch-umount-mount-write pattern
to my growing body of regression tests and try to figure out where the
hang is occurring (I suspect conflicting read_cache_page() calls on
the same page).

Thanks,
Mike
.___________________________________________________________________.
                        Michael A. Halcrow
      Security Software Engineer, IBM Linux Technology Center
GnuPG Fingerprint: 419C 5B1E 948A FA73 A54C 20F5 DB40 8531 6DCA 8769

The last word in this sentence is wrong.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iQEVAwUBRozxFNtAhTFtyodpAQOtnQf/XxS3x8MHw+32AmWs6yqwHYmV+xN/jPUY
Hg4u2sMSzu6KBXL8LG4yaC13XZvI7CNNY5ZatWV4SQRhJXs5nmNbxrxLUYgZrm/q
tlYGrUMRqMX0ajQrjPdgvha4C/8T3nKCy4dAXyTJ/KFiOTjJ8vkABeOpIw9jWjIC
XevokDYwtMUNA4c3408idvNxlLIQX06etVH1bYEeO5plzE/IYGDSYCz3ATntZwKU
gY/igI3+fqouMCattVO4rbnlOGWaOofLseTadgZosS4Ka2dtJuu7XFGbOdYR49Ew
15ituFUPlmV/nmLrWi9cO8DIAojpsDvWvRyc5smr8RaIPrUf6BqRnQ==
=KPen
-----END PGP SIGNATURE-----