Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
On Wed, Sep 17, 2014 at 11:05:06AM +0200, Sebastian Andrzej Siewior wrote: On 2014-09-10 21:29:55 [+0200], Sebastian Andrzej Siewior wrote: the diff of v8…v9 is small: Greg, do you mind taking patches from this series up to [PATCH 05/16]? Nobody complained about those so far and it would keep v10 a little smaller. I have changes to #6 (the omap driver) and need to do some DMA related changes in the following Now applied, thanks. greg k-h -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
On 09/16/2014 11:30 PM, Tony Lindgren wrote: Found one more issue when booting on 2420 n8x0, maybe something to do with runtime PM? To some degree, yes. [4.770507] Internal error: Oops - undefined instruction: 0 [#1] SMP ARM e3a02000mov r2, #0 ee072fbamcr 15, 0, r2, cr7, cr10, {5} e3a01001mov r1, #1 f5d3f000pld [r3] =e1d30f9fldrexb r0, [r3] That ldrexb is part of the xchg() function in serial8250_rpm_get_tx(). So it looks like 2420 n8x0 does not understand ldrexb but the inline assembly decided that it should. This OMAP2420 should be ARM1136 / ARMv6. The ARM1136J(F)-S TRM says for ldrexb: This command is only available from the rev1 (r1p0) release of the ARM1136JF-S processor. So it looks like the CPU should know what to do when this opcode comes around. Regards, Tony Sebastian -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
On 2014-09-10 21:29:55 [+0200], Sebastian Andrzej Siewior wrote: the diff of v8…v9 is small: Greg, do you mind taking patches from this series up to [PATCH 05/16]? Nobody complained about those so far and it would keep v10 a little smaller. I have changes to #6 (the omap driver) and need to do some DMA related changes in the following Sebastian -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
On 09/15/2014 01:50 PM, Sebastian Andrzej Siewior wrote: I changed the restore function the fix the am335x-evm + less freeze. And now we have this. So let me search… IER was 0 by accident. It fixed in TX path. I pushed uart_v10_pre1 which should have it fixed aport from other things… I am going to address the review comments, to split the DMA callbacks as requested and if nobody comes up with something fancy we will have a v10 :) Regards, Tony Sebastian -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
* Sebastian Andrzej Siewior bige...@linutronix.de [140916 05:58]: On 09/15/2014 01:50 PM, Sebastian Andrzej Siewior wrote: I changed the restore function the fix the am335x-evm + less freeze. And now we have this. So let me search… IER was 0 by accident. It fixed in TX path. I pushed uart_v10_pre1 which should have it fixed aport from other things… Just verified that works for my off-idle test case thanks. I am going to address the review comments, to split the DMA callbacks as requested and if nobody comes up with something fancy we will have a v10 :) OK Regards, Tony -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
* Tony Lindgren t...@atomide.com [140916 09:49]: * Sebastian Andrzej Siewior bige...@linutronix.de [140916 05:58]: On 09/15/2014 01:50 PM, Sebastian Andrzej Siewior wrote: I changed the restore function the fix the am335x-evm + less freeze. And now we have this. So let me search… IER was 0 by accident. It fixed in TX path. I pushed uart_v10_pre1 which should have it fixed aport from other things… Just verified that works for my off-idle test case thanks. I am going to address the review comments, to split the DMA callbacks as requested and if nobody comes up with something fancy we will have a v10 :) OK Found one more issue when booting on 2420 n8x0, maybe something to do with runtime PM? Regards, Tony [4.770507] Internal error: Oops - undefined instruction: 0 [#1] SMP ARM [4.777343] Modules linked in: [4.780487] CPU: 0 PID: 1 Comm: init Not tainted 3.17.0-rc5-00211-gc2182d0-dirty #1408 [4.788482] task: c5842b80 ti: c585 task.ti: c585 [4.793945] PC is at serial8250_start_tx+0x124/0x154 [4.798980] LR is at uart_start+0x4c/0x5c [4.803039] pc : [c03d7340]lr : [c03d1ee4]psr: a093 [4.803039] sp : c5851e10 ip : c5851e28 fp : c5851e24 [4.814605] r10: 0007 r9 : r8 : c134d554 [4.819885] r7 : c5cb7800 r6 : 2013 r5 : c134d554 r4 : c134d554 [4.826477] r3 : c134d690 r2 : r1 : 0001 r0 : c134d554 [4.833068] Flags: NzCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user [4.840332] Control: 00c5387d Table: 85cd8000 DAC: 0015 [4.846130] Process init (pid: 1, stack limit = 0xc5850248) [4.851776] Stack: (0xc5851e10 to 0xc5852000) [4.856201] 1e00: c5cb7800 c134d554 c5851e44 c5851e28 [4.864440] 1e20: c03d1ee4 c03d7228 c5842b80 c5b00408 c5cf3c07 c5851e7c c5851e48 [4.872711] 1e40: c03d2da0 c03d1ea4 c5851e6c a013 c03d1fe0 c5cb7800 0007 224c [4.880981] 1e60: c88cc2a0 2250 c5cf3c00 c88ca000 c5851edc c5851e80 c03b9474 c03d2cd4 [4.889251] 1e80: c5851e9c c5cb7ab4 c5cf3c00 c5cb793c c5a5ae00 c585 c0158e1c [4.897491] 1ea0: c5842b80 c007152c c5cb7ad4 c5cb7ad4 c5cb7800 e0cc 0007 c5cb7800 [4.905761] 1ec0: c585 0400 c5a5ae00 c5851f1c c5851ee0 c03b61c0 c03b92d0 [4.914031] 1ee0: 0007 0007 c5cb18c0 c03b92c4 c0b19f20 0007 e0cc [4.922271] 1f00: c5851f78 c5a5ae00 c585 e0cc c5851f44 c5851f20 c03b63c4 c03b60c0 [4.930541] 1f20: c5851f78 c5a5ae00 0007 e0cc c5851f78 0007 c5851f74 c5851f48 [4.938812] 1f40: c0168f38 c03b632c c0185718 c0185688 c5a5ae00 c5a5ae00 [4.947052] 1f60: 0007 e0cc c5851fa4 c5851f78 c0169380 c0168e90 [4.955322] 1f80: e0cc 0001 0004 c000f164 c5851fa8 [4.963592] 1fa0: c000eee0 c0169340 e0cc e0cc 0007 [4.971862] 1fc0: e0cc 0001 0004 a24c 00017504 1000 [4.980102] 1fe0: bef5aa40 bef5aa30 a474 b6e801ec 6010 [4.988372] [c03d7340] (serial8250_start_tx) from [c03d1ee4] (uart_start+0x4c/0x5c) [4.996490] [c03d1ee4] (uart_start) from [c03d2da0] (uart_write+0xd8/0x100) [5.003875] [c03d2da0] (uart_write) from [c03b9474] (n_tty_write+0x1b0/0x510) [5.011474] [c03b9474] (n_tty_write) from [c03b61c0] (tty_write+0x10c/0x26c) [5.018951] [c03b61c0] (tty_write) from [c03b63c4] (redirected_tty_write+0xa4/0xb8) [5.027038] [c03b63c4] (redirected_tty_write) from [c0168f38] (vfs_write+0xb4/0x1bc) [5.035217] [c0168f38] (vfs_write) from [c0169380] (SyS_write+0x4c/0x98) [5.042358] [c0169380] (SyS_write) from [c000eee0] (ret_fast_syscall+0x0/0x48) [5.050018] Code: e3a02000 ee072fba e3a01001 f5d3f000 (e1d30f9f) -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
On 09/13/2014 12:43 AM, Tony Lindgren wrote: * Sebastian Andrzej Siewior bige...@linutronix.de [140910 12:30]: the diff of v8…v9 is small: - rebased on top's of Greg's tty-next branch - fixed #10 where we might have THRE interrupt enabled for longer than needed - re-did register setup in #10. Before this less file could freeze the am335x-evm. FYI, looks like merging in your uart_v9 branch for testing with my hwmod changes now fails to produce RX characters on omap3. The device wakes up just fine to the wake-up interrupt, and then serial8250_handle_irq() gets called, but I'm not seeing anything in the console getting printed out. It's like all the RX characters are getting lost instead of just the first one. The RX characters work fine when the system is running. So it seems there's been some kind of regression since v8? I changed the restore function the fix the am335x-evm + less freeze. And now we have this. So let me search… Regards, Tony Sebastian -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 00/16 v9] omap 8250 based uart + DMA
* Sebastian Andrzej Siewior bige...@linutronix.de [140910 12:30]: the diff of v8…v9 is small: - rebased on top's of Greg's tty-next branch - fixed #10 where we might have THRE interrupt enabled for longer than needed - re-did register setup in #10. Before this less file could freeze the am335x-evm. FYI, looks like merging in your uart_v9 branch for testing with my hwmod changes now fails to produce RX characters on omap3. The device wakes up just fine to the wake-up interrupt, and then serial8250_handle_irq() gets called, but I'm not seeing anything in the console getting printed out. It's like all the RX characters are getting lost instead of just the first one. The RX characters work fine when the system is running. So it seems there's been some kind of regression since v8? Regards, Tony -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 00/16 v9] omap 8250 based uart + DMA
the diff of v8…v9 is small: - rebased on top's of Greg's tty-next branch - fixed #10 where we might have THRE interrupt enabled for longer than needed - re-did register setup in #10. Before this less file could freeze the am335x-evm. Sebastian -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html