From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dimitar Dimitrov Subject: Re: DSS: kernel p-android-3.0 hangs when omapdss.debug is enabled. Date: Sun, 2 Sep 2012 22:02:10 +0300 Message-ID: <201209022202.10938.dinuxbg@gmail.com> References: <201209021935.11897.dinuxbg@gmail.com> <50439868.9000700@gmail.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from mail-bk0-f46.google.com ([209.85.214.46]:62688 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754417Ab2IBTAq (ORCPT ); Sun, 2 Sep 2012 15:00:46 -0400 Received: by bkwj10 with SMTP id j10so1900442bkw.19 for ; Sun, 02 Sep 2012 12:00:45 -0700 (PDT) In-Reply-To: <50439868.9000700@gmail.com> Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: Chris Hoffmann Cc: linux-omap@vger.kernel.org On Sunday 02 September 2012 20:33:28 Chris Hoffmann wrote: > On 09/02/2012 06:35 PM, Dimitar Dimitrov wrote: > > On Sunday 02 September 2012 15:25:09 Christian Hoffmann wrote: > >> Hi, > >> > >> we're currently porting a 2.6.35 android kernel for a custom omap4430 > >> board (aka nook tablet) to version 3.0.x. We're experiencing random > >> hangs with no trace in the dmesg/last_kmsg or on serial console. Being > >> desperate to find the cause, I tried to put omapdss.debug=y but > >> unfortunately now the kernel hangs on boot. Putting on lock debugging, > >> I can see the following in last_kmsg: > >> ... > >> [ 17.161529] omapdss MANAGER: configure_manager(2) > >> [ 17.166625] omapdss MANAGER: configure manager wbc->shadow_dirty = 0 > >> [ 17.173370] omapdss DISPC: GO LCD2 > >> [ 17.177368] omapdss MANAGER: configure manager wbc->shadow_dirty = 0 > >> [ 17.184143] omapdss MANAGER: configure manager wbc->shadow_dirty = 0 > >> [ 17.191192] omapdss MANAGER: configure manager wbc->shadow_dirty = 0 > >> [ 37.691680] BUG: spinlock lockup on CPU#0, BootAnimation/276, > >> e5fbfd74 [ 37.699035] Backtrace: > >> [ 37.701965] [] (dump_backtrace+0x0/0x114) from > >> [] (dump_stack+0x20/0x24) > >> [ 37.710998] r6:00000000 r5:e5438000 r4:e5fbfd74 r3:00000002 > >> [ 37.717651] [] (dump_stack+0x0/0x24) from [] > >> (do_raw_spin_lock+0x11c/0x180) > >> [ 37.727081] [] (do_raw_spin_lock+0x0/0x180) from > >> [] (_raw_spin_lock_irqsave+0x74/0x80) > >> [ 37.737426] r9:e5438000 r8:00040000 r7:00000002 r6:c009f18c > >> r5:280b0193 [ 37.744995] r4:e5fbfd74 > >> [ 37.748107] [] (_raw_spin_lock_irqsave+0x0/0x80) from > >> [] (complete+0x28/0x68) > >> [ 37.757690] r6:e5fbfd70 r5:00040080 r4:e5fbfd74 > >> [ 37.763031] [] (complete+0x0/0x68) from [] > >> (dispc_irq_wait_handler.27590+0x18/0x1c) > >> [ 37.773162] r6:c0e5850c r5:00040080 r4:e5439d98 > >> [ 37.778594] [] (dispc_irq_wait_handler.27590+0x0/0x1c) > >> from [] (omap_dispc_irq_handler+0xb0/0x220) > >> [ 37.790130] [] (omap_dispc_irq_handler+0x0/0x220) from > >> [] (handle_irq_event_percpu+0x84/0x2dc) > >> [ 37.801300] [] (handle_irq_event_percpu+0x0/0x2dc) from > >> [] (handle_irq_event+0x4c/0x6c) > >> [ 37.811737] [] (handle_irq_event+0x0/0x6c) from > >> [] (handle_fasteoi_irq+0xcc/0x174) > >> [ 37.821777] r6:e5438000 r5:c08a2cac r4:c08a2c60 r3:00000000 > >> [ 37.828338] [] (handle_fasteoi_irq+0x0/0x174) from > >> [] (generic_handle_irq+0x40/0x54) > >> [ 37.838562] r5:c0061158 r4:c08b4260 > >> [ 37.842681] [] (generic_handle_irq+0x0/0x54) from > >> [] (asm_do_IRQ+0x5c/0xbc) > >> [ 37.852111] [] (asm_do_IRQ+0x0/0xbc) from [] > >> (__irq_svc+0x4c/0xec) > >> ... > >> > >> Full dmesg is here: > >> https://dl.dropbox.com/u/21820416/crash-dss-debug.dmesg > >> > >> Kernel source is very close to p-android-3.0. It can be found here: > >> https://github.com/Kuzma30/kernel3NookTablet/tree/jellybean > >> Used config: cyanogenmod_acclaim_defconfig. > >> > >> We have made some slight modifications to dss (for DSI2/lcd2/vsync2). > >> For quick comparison, I put diff off drivers/video/omap2 here: > >> https://dl.dropbox.com/u/21820416/dss.diff > >> > >> Please also note that we have a "boxer" as display > >> (https://github.com/Kuzma30/kernel3NookTablet/blob/jellybean/drivers/vid > >> eo/ omap2/displays/panel-boxer.c). > >> > >> Kind regards, > >> Chris > >> -- > >> To unsubscribe from this list: send the line "unsubscribe linux-omap" in > >> the body of a message to majordomo@vger.kernel.org > >> More majordomo info at http://vger.kernel.org/majordomo-info.html > > > > Could you try the following v3.0 patch: > > http://review.omapzoom.org/#/c/26235/ > > > > Regards, > > Dimitar > > Hi, > > Thank you, very much. This indeed makes the omapdss.debug=y work. I had > to "rebase" the change a bit as we also have this one: > "OMAPFB: Select correct VSYNC IRQ based on DSS channel" in place. > > Thanks again! > > Rgds, > Chris This bug is also present in current linux-omap git tree. I'll post a rebased patch.