From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: 4.16 OMAP serial transmit corruption? Date: Tue, 17 Apr 2018 17:57:48 -0700 Message-ID: <20180418005748.GE5671@atomide.com> References: <20180416151732.GU16141@n2100.armlinux.org.uk> <20180416154545.GA5671@atomide.com> <413b774b-d19f-3221-44d5-7992d3b8757f@ti.com> <20180417173134.GD5671@atomide.com> <20180417221029.GL10990@n2100.armlinux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20180417221029.GL10990@n2100.armlinux.org.uk> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=m.gmane.org@lists.infradead.org To: Russell King - ARM Linux Cc: Greg Kroah-Hartman , linux-omap@vger.kernel.org, Vignesh R , linux-arm-kernel@lists.infradead.org, linux-serial@vger.kernel.org List-Id: linux-omap@vger.kernel.org * Russell King - ARM Linux [180417 22:12]: > On Tue, Apr 17, 2018 at 10:31:35AM -0700, Tony Lindgren wrote: > > * Vignesh R [180417 09:21]: > > > On Monday 16 April 2018 09:15 PM, Tony Lindgren wrote: > > > > Also, I'm seeing an issue where the UARTs won't idle on init > > > > with 8250_omap driver if connected to the wl12xx bluetooth port > > > > unless I write some data to the port first. It does not seem > > > > to be related to the rts/cts lines being wired as I've tested > > > > muxing them out of the way. > > > > > > If this instance of UART is using DMA then it might be due an errata > > > worked around in AM33/AM43/DRA7: > > > https://patchwork.kernel.org/patch/6784331/ > > > > It sure sounds similar but UART_ERRATA_CLOCK_DISABLE does not > > seem to help with the reset. Also disabling DMA does not seem > > to help. So far the only way to clear it seems to be to write > > a character (TX) on the device. Then things work just fine > > even without UART_ERRATA_CLOCK_DISABLE set. I'll try to debug > > this more at some point. > > Classic case of thread hijack. So, Tony's idle problem gets more > attention on _my_ thread than _my_ issue about TX corruption, yea, > that's fair... Come on guys, what about my problem, which is the > subject of this thread? Just trying to brainstorm what all can go wrong still :) OK so it's not DMA then. > I don't have CONFIG_SERIAL_8250_DMA set, so DMA can't be the issue. Is this happening also with v4.16 or with v4.17-rc1? So you just edit something in vi and and on exit it happens? Not happening here for me.. Regards, Tony