From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vineet Gupta Subject: Re: ipc-msg broken again on 3.11-rc7? (was Re: linux-next: Tree for Jun 21 [ BROKEN ipc/ipc-msg ]) Date: Thu, 29 Aug 2013 07:21:49 +0000 Message-ID: References: <1372192414.1888.8.camel@buesod1.americas.hpqcorp.net> <1372202983.1888.22.camel@buesod1.americas.hpqcorp.net> <521DE5D7.4040305@synopsys.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Return-path: Content-Language: en-US Sender: owner-linux-mm@kvack.org To: "sedat.dilek@gmail.com" Cc: Davidlohr Bueso , linux-next , LKML , Stephen Rothwell , Andrew Morton , linux-mm , Andi Kleen , Rik van Riel , Manfred Spraul , Jonathan Gonzalez List-Id: linux-next.vger.kernel.org On 08/29/2013 08:34 AM, Sedat Dilek wrote:=0A= > On Wed, Aug 28, 2013 at 1:58 PM, Vineet Gupta=0A= > wrote:=0A= >> Hi David,=0A= >>=0A= >> On 06/26/2013 04:59 AM, Davidlohr Bueso wrote:=0A= >>> On Tue, 2013-06-25 at 23:41 +0200, Sedat Dilek wrote:=0A= >>>> On Tue, Jun 25, 2013 at 10:33 PM, Davidlohr Bueso=0A= >>>> wrote:=0A= >>>>> On Tue, 2013-06-25 at 18:10 +0200, Sedat Dilek wrote:=0A= >>>>> [...]=0A= >>>>>=0A= >>>>>> I did some more testing with Linux-Testing-Project (release:=0A= >>>>>> ltp-full-20130503) and next-20130624 (Monday) which has still the=0A= >>>>>> issue, here.=0A= >>>>>>=0A= >>>>>> If I revert the mentioned two commits from my local=0A= >>>>>> revert-ipc-next20130624-5089fd1c6a6a-ab9efc2d0db5 GIT repo, everythi= ng=0A= >>>>>> is fine.=0A= >>>>>>=0A= >>>>>> I have tested the LTP ***IPC*** and ***SYSCALLS*** testcases.=0A= >>>>>>=0A= >>>>>> root# ./runltp -f ipc=0A= >>>>>>=0A= >>>>>> root# ./runltp -f syscalls=0A= >>>>> These are nice test cases!=0A= >>>>>=0A= >>>>> So I was able to reproduce the issue with LTP and manually running=0A= >>>>> msgctl08. We seemed to be racing at find_msg(), so take to q_perm loc= k=0A= >>>>> before calling it. The following changes fixes the issue and passes a= ll=0A= >>>>> 'runltp -f syscall' tests, could you give it a try?=0A= >>>>>=0A= >>>> Cool, that fixes the issues here.=0A= >>>>=0A= >>>> Building with fakeroot & make deb-pkg is now OK, again.=0A= >>>>=0A= >>>> The syscalls/msgctl08 test-case ran successfully!=0A= >>> Andrew, could you pick this one up? I've made the patch on top of=0A= >>> 3.10.0-rc7-next-20130625=0A= >> LTP msgctl08 hangs on 3.11-rc7 (ARC port) with some of my local changes.= I=0A= >> bisected it, sigh... didn't look at this thread earlier :-( and landed i= nto this.=0A= >>=0A= >> ------------->8------------------------------------=0A= >> 3dd1f784ed6603d7ab1043e51e6371235edf2313 is the first bad commit=0A= >> commit 3dd1f784ed6603d7ab1043e51e6371235edf2313=0A= >> Author: Davidlohr Bueso =0A= >> Date: Mon Jul 8 16:01:17 2013 -0700=0A= >>=0A= >> ipc,msg: shorten critical region in msgsnd=0A= >>=0A= >> do_msgsnd() is another function that does too many things with the i= pc=0A= >> object lock acquired. Take it only when needed when actually updati= ng=0A= >> msq.=0A= >> ------------->8------------------------------------=0A= >>=0A= >> If I revert 3dd1f784ed66 and 9ad66ae "ipc: remove unused functions" - th= e test=0A= >> passes. I can confirm that linux-next also has the issue (didn't try the= revert=0A= >> there though).=0A= >>=0A= >> 1. arc 3.11-rc7 config attached (UP + PREEMPT)=0A= >> 2. dmesg prints "msgmni has been set to 479"=0A= >> 3. LTP output (this is slightly dated source, so prints might vary)=0A= >>=0A= >> ------------->8------------------------------------=0A= >> <<>>=0A= >> tag=3Dmsgctl08 stime=3D1377689180=0A= >> cmdline=3D"msgctl08"=0A= >> contacts=3D""=0A= >> analysis=3Dexit=0A= >> initiation_status=3D"ok"=0A= >> <<>>=0A= >> ------------->8-------- hung here ------------------=0A= >>=0A= >>=0A= >> Let me know if you need more data/test help.=0A= >>=0A= > Cannot say much to your constellation as I had the issue on x86-64 and=0A= > Linux-next.=0A= > But I have just seen a post-v3.11-rc7 IPC-fix in [1].=0A= >=0A= > I have here a v3.11-rc7 kernel with drm-intel-nightly on top... did not r= un LTP.=0A= =0A= Not sure what you mean - I'd posted that Im seeing the issue on ARC Linux (= an FPGA=0A= board) 3.11-rc7 as well as linux-next of yesterday.=0A= =0A= > Which LTP release do you use?=0A= =0A= The LTP build I generally use is from a 2007 based sources (lazy me). Howev= er I=0A= knew this would come up so before posting, I'd built the latest from buildr= oot and=0A= ran the msgctl08 from there standalone and it did the same thing.=0A= =0A= > Might be good to attach your kernel-config for followers?=0A= =0A= It was already there in my orig msg - you probably missed it.=0A= =0A= > [1] http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit= /?id=3D368ae537e056acd3f751fa276f48423f06803922=0A= =0A= I tried linux-next of today, same deal - msgctl08 still hangs.=0A= =0A= -Vineet=0A= -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org