On Tue, 2013-01-15 at 03:29 -0500, CAI Qian wrote: > > Hmm David, a bit confused here. > b01af457 (the offender) was merged in v3.5-rc1.
This fixed a theoretical bug, at the cost of breaking the driver for
real hardware.
> b26623da (the revert) was merged in v3.7-rc7.
The simple and obvious fix for stable, which obviously re-introduced the
original theoretical bug. This is why you see context differences :)
> a9dbe40f (this commit marked for stable) was merged in v3.8-rc1.
This was my own fix, which fixed everything. But Dave had decided *not*
to do this in the stable series; just to revert the original patch.
Perhaps Dave has changed his mind now, and my fix can be backported to
stable. But we still haven't seen any evidence that the original bug
fixed by commit b01af457 was really causing anyone any problems — so he
may well prefer to leave things as they are.
Your version of the patch is technically correct and does the right
thing... *if* that's actually what we want to do in -stable.
--
Sent with MeeGo's ActiveSync support.
David Woodhouse Open Source Technology Centre
[email protected] Intel Corporation
smime.p7s
Description: S/MIME cryptographic signature
