Return-Path: Date: Fri, 10 Feb 2012 10:09:43 +0200 From: Emeltchenko Andrei To: Ulisses Furquim Cc: Marcel Holtmann , linux-bluetooth@vger.kernel.org Subject: Re: [RFCv3 02/16] Bluetooth: Revert to mutexes from RCU list Message-ID: <20120210080941.GA28197@aemeltch-MOBL1> References: <1328797057-26331-1-git-send-email-Andrei.Emeltchenko.news@gmail.com> <1328797057-26331-3-git-send-email-Andrei.Emeltchenko.news@gmail.com> <1328797634.28848.21.camel@aeonflux> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 In-Reply-To: List-ID: Hi Ulisses, On Thu, Feb 09, 2012 at 03:54:44PM -0200, Ulisses Furquim wrote: > Hi Andrei, > > On Thu, Feb 9, 2012 at 12:27 PM, Marcel Holtmann wrote: > > Hi Andrei, > > > >> Usage of RCU list looks not reasonalbe for a number of reasons: > >> our code sleep and we have to use socket spinlocks, some parts > >> of code are updaters thus we need to use mutexes anyway. > >> > >> Signed-off-by: Andrei Emeltchenko > >> --- > >> ?net/bluetooth/l2cap_core.c | ?108 ++++++++++++++++++++++---------------------- > >> ?1 files changed, 54 insertions(+), 54 deletions(-) > > > > I am fine with this, but I wanna get an ack from Ulisses as well. > > Please, refer to my previous review because it seems you haven't > changed anything at all. For instance, l2cap_disconnect_rsp() is still those changes are in other patch: [RFCv3 09/16] Bluetooth: Use chan lock in L2CAP sig commands I was thinking that this patch change RCU to mutex chan list locks and the changes which change socket lock to chan lock are different. > missing an unlock of chan_lock and I see no changes to l2cap_sock.c > which are needed. Moreover, you need to use the unlocked versions of > the chan lookup functions and check if we still need the locked ones. I will check need for locking function but at least compiler did not warn me about unused function. Best regards Andrei Emeltchenko