From mboxrd@z Thu Jan 1 00:00:00 1970 Subject: Re: 2.6.2-mm1 aka "Geriatric Wombat" DIO read race still fails From: Daniel McNeil In-Reply-To: <200402051558.08927.pbadari@us.ibm.com> References: <20040205014405.5a2cf529.akpm@osdl.org> <1076023899.7182.97.camel@ibm-c.pdx.osdl.net> <200402051558.08927.pbadari@us.ibm.com> Content-Type: text/plain Message-Id: <1076107043.7182.146.camel@ibm-c.pdx.osdl.net> Mime-Version: 1.0 Date: 06 Feb 2004 14:37:24 -0800 Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: Badari Pulavarty Cc: Andrew Morton , Linux Kernel Mailing List , linux-mm@kvack.org, "linux-aio@kvack.org" List-ID: I patched 2.6.2-mm1 with the wait_on_buffer(bh) in __block_write_full_page() and my tests ran overnight and though today without any errors. I'm looking to see if there is any way to move the end_page_writeback() for this locked bh case to the end_buffer_write_sync(). Kind of tricky... Daniel On Thu, 2004-02-05 at 15:58, Badari Pulavarty wrote: > On Thursday 05 February 2004 03:31 pm, Daniel McNeil wrote: > > Andrew, > > > > I tested 2.6.2-mm1 on an 8-proc running 6 copies of the read_under > > test and all 6 read_under tests saw uninitialized data in less than 5 > > minutes. :( > > > > Daniel > > Daniel, > > Same here... Just FYI, I am running with your original patch and > not failed so far (2 hours..) Normally, I see the problem in 15 min or so. > > Thanks, > Badari -- 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: aart@kvack.org