Re: [PATCH 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag
* Vikram Pandita vikram.pand...@ti.com [091116 15:00]: OMAP3630 and OMAP4430 UART IP blocks have a restriction wrt RX FIFO. Empty RX fifo read causes an abort. OMAP1/2/3 do not have this restriction. So interoduce a flag in 8250 driver: UPF_NO_EMPTY_FIFO_READ If this flag is specified for an 8250 port, then first check if rx fifo has contents, and only then proceed to read the fifo. The change affects only boards passing this flag in port data and hence is non-disruptive for other boards. First reported/fixed on omap4430 by Shilimkar, Santosh This should go to linux-ser...@vger.kernel.org and Alan Cox. Regards, Tony Signed-off-by: Vikram Pandita vikram.pand...@ti.com Cc: Shilimkar, Santosh santosh.shilim...@ti.com --- drivers/serial/8250.c | 30 +- include/linux/serial_core.h |1 + 2 files changed, 26 insertions(+), 5 deletions(-) diff --git a/drivers/serial/8250.c b/drivers/serial/8250.c index 737b4c9..03c890e 100644 --- a/drivers/serial/8250.c +++ b/drivers/serial/8250.c @@ -1245,7 +1245,11 @@ static void autoconfig(struct uart_8250_port *up, unsigned int probeflags) #endif serial_outp(up, UART_MCR, save_mcr); serial8250_clear_fifos(up); - serial_in(up, UART_RX); + if (up-port.flags UPF_NO_EMPTY_FIFO_READ) { + if (serial_inp(up, UART_LSR) UART_LSR_DR) + serial_in(up, UART_RX); + } else + serial_in(up, UART_RX); if (up-capabilities UART_CAP_UUE) serial_outp(up, UART_IER, UART_IER_UUE); else @@ -1289,7 +1293,11 @@ static void autoconfig_irq(struct uart_8250_port *up) } serial_outp(up, UART_IER, 0x0f);/* enable all intrs */ (void)serial_inp(up, UART_LSR); - (void)serial_inp(up, UART_RX); + if (up-port.flags UPF_NO_EMPTY_FIFO_READ) { + if (serial_inp(up, UART_LSR) UART_LSR_DR) + (void)serial_inp(up, UART_RX); + } else + (void)serial_inp(up, UART_RX); (void)serial_inp(up, UART_IIR); (void)serial_inp(up, UART_MSR); serial_outp(up, UART_TX, 0xFF); @@ -1984,7 +1992,11 @@ static int serial8250_startup(struct uart_port *port) * Clear the interrupt registers. */ (void) serial_inp(up, UART_LSR); - (void) serial_inp(up, UART_RX); + if (up-port.flags UPF_NO_EMPTY_FIFO_READ) { + if (serial_inp(up, UART_LSR) UART_LSR_DR) + (void) serial_inp(up, UART_RX); + } else + (void) serial_inp(up, UART_RX); (void) serial_inp(up, UART_IIR); (void) serial_inp(up, UART_MSR); @@ -2141,7 +2153,11 @@ dont_test_tx_en: * routines or the previous session. */ serial_inp(up, UART_LSR); - serial_inp(up, UART_RX); + if (up-port.flags UPF_NO_EMPTY_FIFO_READ) { + if (serial_inp(up, UART_LSR) UART_LSR_DR) + serial_inp(up, UART_RX); + } else + serial_inp(up, UART_RX); serial_inp(up, UART_IIR); serial_inp(up, UART_MSR); up-lsr_saved_flags = 0; @@ -2207,7 +2223,11 @@ static void serial8250_shutdown(struct uart_port *port) * Read data port to reset things, and then unlink from * the IRQ chain. */ - (void) serial_in(up, UART_RX); + if (up-port.flags UPF_NO_EMPTY_FIFO_READ) { + if (serial_inp(up, UART_LSR) UART_LSR_DR) + (void) serial_inp(up, UART_RX); + } else + (void) serial_in(up, UART_RX); del_timer_sync(up-timer); up-timer.function = serial8250_timeout; diff --git a/include/linux/serial_core.h b/include/linux/serial_core.h index db532ce..0d65bb3 100644 --- a/include/linux/serial_core.h +++ b/include/linux/serial_core.h @@ -308,6 +308,7 @@ struct uart_port { #define UPF_SPD_WARP ((__force upf_t) (0x1010)) #define UPF_SKIP_TEST((__force upf_t) (1 6)) #define UPF_AUTO_IRQ ((__force upf_t) (1 7)) +#define UPF_NO_EMPTY_FIFO_READ ((__force upf_t) (1 8)) #define UPF_HARDPPS_CD ((__force upf_t) (1 11)) #define UPF_LOW_LATENCY ((__force upf_t) (1 13)) #define UPF_BUGGY_UART ((__force upf_t) (1 14)) -- 1.6.5.1.69.g36942 -- 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 -- 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 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag
Tony From: linux-omap-ow...@vger.kernel.org [mailto:linux-omap-ow...@vger.kernel.org] On Behalf Of Tony Lindgren Sent: Tuesday, November 17, 2009 11:06 AM To: Pandita, Vikram Cc: linux-omap@vger.kernel.org; Shilimkar, Santosh Subject: Re: [PATCH 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag * Vikram Pandita vikram.pand...@ti.com [091116 15:00]: OMAP3630 and OMAP4430 UART IP blocks have a restriction wrt RX FIFO. Empty RX fifo read causes an abort. OMAP1/2/3 do not have this restriction. So interoduce a flag in 8250 driver: UPF_NO_EMPTY_FIFO_READ If this flag is specified for an 8250 port, then first check if rx fifo has contents, and only then proceed to read the fifo. The change affects only boards passing this flag in port data and hence is non-disruptive for other boards. First reported/fixed on omap4430 by Shilimkar, Santosh This should go to linux-ser...@vger.kernel.org and Alan Cox. Patches were posted to l-o for review. Next step was to get it to the right subsystem. I thought Alan Cox stepped down from the maintainership of serial/tty subsystem. Not sure if Andrew Morton is taking these patches to MM tree for now. Any idea? Regards, Tony snip -- 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 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag
* Pandita, Vikram vikram.pand...@ti.com [091117 09:39]: Tony From: linux-omap-ow...@vger.kernel.org [mailto:linux-omap-ow...@vger.kernel.org] On Behalf Of Tony Lindgren Sent: Tuesday, November 17, 2009 11:06 AM To: Pandita, Vikram Cc: linux-omap@vger.kernel.org; Shilimkar, Santosh Subject: Re: [PATCH 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag * Vikram Pandita vikram.pand...@ti.com [091116 15:00]: OMAP3630 and OMAP4430 UART IP blocks have a restriction wrt RX FIFO. Empty RX fifo read causes an abort. OMAP1/2/3 do not have this restriction. So interoduce a flag in 8250 driver: UPF_NO_EMPTY_FIFO_READ If this flag is specified for an 8250 port, then first check if rx fifo has contents, and only then proceed to read the fifo. The change affects only boards passing this flag in port data and hence is non-disruptive for other boards. First reported/fixed on omap4430 by Shilimkar, Santosh This should go to linux-ser...@vger.kernel.org and Alan Cox. Patches were posted to l-o for review. Next step was to get it to the right subsystem. I thought Alan Cox stepped down from the maintainership of serial/tty subsystem. Not sure if Andrew Morton is taking these patches to MM tree for now. Any idea? OK, no idea who's dealing with 8250 now :) 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 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag
On Tue, 17 Nov 2009 15:16:55 -0600 Vikram Pandita vikram.pand...@ti.com wrote: OMAP3630 and OMAP4430 UART IP blocks have a restriction wrt RX FIFO. Empty RX fifo read causes an abort. OMAP1/2/3 do not have this restriction. Nothing wrong with the requirement but I think it would keep 8250.c a lot cleaner if you implemented that in the serial_in method for the OMAP3630/4430. It would also mean anyone adding a UART_RX or any cases you miss don't break in future This is especially true now as you can pass a private serial_in/serial_out method in the port register function. Care to submit an alternate patch doing it that way, or see any reason for not ? Alan -- 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 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag
-Original Message- From: Alan Cox [mailto:a...@lxorguk.ukuu.org.uk] Sent: Tuesday, November 17, 2009 3:26 PM To: Pandita, Vikram Cc: linux-ser...@vger.kernel.org; a...@linux-foundation.org; linux-omap@vger.kernel.org; Pandita, Vikram; Shilimkar, Santosh Subject: Re: [PATCH 1/2] serial: 8250: add UPF_NO_EMPTY_FIFO_READ flag On Tue, 17 Nov 2009 15:16:55 -0600 Vikram Pandita vikram.pand...@ti.com wrote: OMAP3630 and OMAP4430 UART IP blocks have a restriction wrt RX FIFO. Empty RX fifo read causes an abort. OMAP1/2/3 do not have this restriction. Nothing wrong with the requirement but I think it would keep 8250.c a lot cleaner if you implemented that in the serial_in method for the OMAP3630/4430. It would also mean anyone adding a UART_RX or any cases you miss don't break in future Agree. Was contemplating this change in beginning, but lost that train of thought. Thx for your review time. This is especially true now as you can pass a private serial_in/serial_out method in the port register function. Our requirement is to over-right the serial_in for now. Is it fine to just add serial_in over-load or do you suggest both in/out? Care to submit an alternate patch doing it that way, or see any reason for not ? Working on it right away. See my patch in an hours time. Alan -- 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