From: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Organization: Tele-Info-System, Poznan, PL
To: Tony Lindgren <tony@atomide.com>
Subject: Re: [PATCH v3] OMAP: DMA: Fix omapfb/lcdc on OMAP1510 broken when PM set
Date: Tue, 10 Nov 2009 03:50:19 +0100
User-Agent: KMail/1.9.10
Cc: linux-omap@vger.kernel.org, linux-fbdev-devel@lists.sourceforge.net,
       linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org
References: <200911051951.32020.jkrzyszt@tis.icnet.pl> <20091109211424.GF23952@atomide.com>
In-Reply-To: <20091109211424.GF23952@atomide.com>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 8BIT
Content-Disposition: inline
Message-Id: <200911100350.22368.jkrzyszt@tis.icnet.pl>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 2627
Lines: 70

Monday 09 November 2009 22:14:24 Tony Lindgren napisał(a):
> * Janusz Krzysztofik <jkrzyszt@tis.icnet.pl> [091105 10:52]:
> > With CONFIG_PM=y, the omapfb/lcdc device on Amstrad Delta, after
> > initially starting correctly, breaks with the following error messages:
> >
> > omapfb omapfb: resetting (status 0xffffff96,reset count 1)
> > ...
> > omapfb omapfb: resetting (status 0xffffff96,reset count 100)
> > omapfb omapfb: too many reset attempts, giving up.
> >
> > Looking closer at this I have found that it had been broken almost 2
> > years ago with commit 2418996e3b100114edb2ae110d5d4acb928909d2, PM fixes
> > for OMAP1.
> >
> > The definite reason for broken omapfb/lcdc behavoiur in PM mode
> > appeared to be ARM_IDLECT1:IDLIF_ARM (bit 6) put into idle regardless of
> > LCD DMA possibly running. The bit were set based on return value of the
> > omap_dma_running() function that did not check for dedicated LCD DMA
> > channel status. The patch below fixes this.
> >
> > Created against linux-2.6.32-rc6
> >
> > Tested on Amstrad Delta
> >
> > Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
> >
> > ---
> >
> > Tuesday 03 November 2009 21:10:48 Tony Lindgren wrote:
> > > * Janusz Krzysztofik <jkrzyszt@tis.icnet.pl> [091103 12:05]:
> > > > Sorry, I've not checked for new mail before posting this one and
> > > > missed your acceptance for my idea of fixing all omap1510, not only
> > > > ams_delta. If there are no more comments, I'll submit v3 with
> > > > cpu_is_omap1510() replacing machine_is_ams_delta().
> > >
> > > Sounds good to me.
> > >
> > > Tony
> >
> > --- linux-2.6.32-rc6/arch/arm/plat-omap/dma.c.orig	2009-11-03 20:37:49.000000000 +0100
> > +++ linux-2.6.32-rc6/arch/arm/plat-omap/dma.c	2009-11-05 19:30:39.000000000 +0100
> > @@ -1108,6 +1108,14 @@ int omap_dma_running(void)
> >  {
> >  	int lch;
> >
> > +	/*
> > +	 * On OMAP1510, internal LCD controller will start the transfer
> > +	 * when it gets enabled, so assume DMA running if LCD enabled.
> > +	 */
> > +	if (cpu_is_omap1510())
> > +		if (omap_readw(0xfffec000 + 0x00) & (1 << 0))
> > +			return 1;
> > +
> >  	/* Check if LCD DMA is running */
> >  	if (cpu_is_omap16xx())
> >  		if (omap_readw(OMAP1610_DMA_LCD_CCR) & OMAP_DMA_CCR_EN)
>
> Can you please use some define rather than the 0xfffec000 register
> above?

Tony,
Sure. Expect a follow up very soon.

Thanks,
Janusz
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/