| Message ID | 1427890824-1599-1-git-send-email-davidcdueck@googlemail.com |
|---|---|
| State | Accepted |
| Delegated to: | Jagannadha Sutradharudu Teki |
| Headers | show |
On 1 April 2015 at 17:50, David Dueck <davidcdueck@googlemail.com> wrote: > The timeout value is never reset during the transfer. This means that when > transferring more data we eventually trigger the timeout. > > This was reported on the mailing list: > "Spansion SPI flash read timeout with AM335x" > > Signed-off-by: David Dueck <davidcdueck@googlemail.com> > CC: Tom Rini <trini@konsulko.com> > CC: Jagannadh Teki <jagannadh.teki@gmail.com> > CC: Stefan Roese <sr@denx.de> > CC: Andy Pont <andy.pont@sdcsystems.com> > --- > Changes since v1: > - fix style issue > - fix CC line > > drivers/spi/omap3_spi.c | 20 ++++++++++++-------- > 1 file changed, 12 insertions(+), 8 deletions(-) Any Tested-by? > > diff --git a/drivers/spi/omap3_spi.c b/drivers/spi/omap3_spi.c > index 651e46e..85f9e85 100644 > --- a/drivers/spi/omap3_spi.c > +++ b/drivers/spi/omap3_spi.c > @@ -20,7 +20,7 @@ > #include <asm/io.h> > #include "omap3_spi.h" > > -#define SPI_WAIT_TIMEOUT 3000000 > +#define SPI_WAIT_TIMEOUT 10 > > static void spi_reset(struct omap3_spi_slave *ds) > { > @@ -227,7 +227,7 @@ int omap3_spi_write(struct spi_slave *slave, unsigned int len, const void *txp, > { > struct omap3_spi_slave *ds = to_omap3_spi(slave); > int i; > - int timeout = SPI_WAIT_TIMEOUT; > + ulong start; > int chconf = readl(&ds->regs->channel[ds->slave.cs].chconf); > > /* Enable the channel */ > @@ -241,9 +241,10 @@ int omap3_spi_write(struct spi_slave *slave, unsigned int len, const void *txp, > > for (i = 0; i < len; i++) { > /* wait till TX register is empty (TXS == 1) */ > + start = get_timer(0); > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & > OMAP3_MCSPI_CHSTAT_TXS)) { > - if (--timeout <= 0) { > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { > printf("SPI TXS timed out, status=0x%08x\n", > readl(&ds->regs->channel[ds->slave.cs].chstat)); > return -1; > @@ -280,7 +281,7 @@ int omap3_spi_read(struct spi_slave *slave, unsigned int len, void *rxp, > { > struct omap3_spi_slave *ds = to_omap3_spi(slave); > int i; > - int timeout = SPI_WAIT_TIMEOUT; > + ulong start; > int chconf = readl(&ds->regs->channel[ds->slave.cs].chconf); > > /* Enable the channel */ > @@ -295,10 +296,11 @@ int omap3_spi_read(struct spi_slave *slave, unsigned int len, void *rxp, > writel(0, &ds->regs->channel[ds->slave.cs].tx); > > for (i = 0; i < len; i++) { > + start = get_timer(0); > /* Wait till RX register contains data (RXS == 1) */ > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & > OMAP3_MCSPI_CHSTAT_RXS)) { > - if (--timeout <= 0) { > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { > printf("SPI RXS timed out, status=0x%08x\n", > readl(&ds->regs->channel[ds->slave.cs].chstat)); > return -1; > @@ -332,7 +334,7 @@ int omap3_spi_txrx(struct spi_slave *slave, unsigned int len, > const void *txp, void *rxp, unsigned long flags) > { > struct omap3_spi_slave *ds = to_omap3_spi(slave); > - int timeout = SPI_WAIT_TIMEOUT; > + ulong start; > int chconf = readl(&ds->regs->channel[ds->slave.cs].chconf); > int irqstatus = readl(&ds->regs->irqstatus); > int i=0; > @@ -350,9 +352,10 @@ int omap3_spi_txrx(struct spi_slave *slave, unsigned int len, > for (i=0; i < len; i++){ > /* Write: wait for TX empty (TXS == 1)*/ > irqstatus |= (1<< (4*(ds->slave.bus))); > + start = get_timer(0); > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & > OMAP3_MCSPI_CHSTAT_TXS)) { > - if (--timeout <= 0) { > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { > printf("SPI TXS timed out, status=0x%08x\n", > readl(&ds->regs->channel[ds->slave.cs].chstat)); > return -1; > @@ -368,9 +371,10 @@ int omap3_spi_txrx(struct spi_slave *slave, unsigned int len, > writel(((u8 *)txp)[i], tx); > > /*Read: wait for RX containing data (RXS == 1)*/ > + start = get_timer(0); > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & > OMAP3_MCSPI_CHSTAT_RXS)) { > - if (--timeout <= 0) { > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { > printf("SPI RXS timed out, status=0x%08x\n", > readl(&ds->regs->channel[ds->slave.cs].chstat)); > return -1; > -- > 2.3.4 >
Hi David, <snipped for brevity> > for (i = 0; i < len; i++) { > /* wait till TX register is empty (TXS == 1) */ > + start = get_timer(0); > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & > OMAP3_MCSPI_CHSTAT_TXS)) { > - if (--timeout <= 0) { > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { > printf("SPI TXS timed out, status=0x%08x\n", > readl(&ds->regs->channel[ds- > >slave.cs].chstat)); > return -1; I have a couple of questions... Firstly, when in SPL is there access to the get_timer() function? Secondly, when using Falcon mode to load Linux directly from SPI (Falcon mode) then we want to maximise the throughput and save every CPU cycle we possibly can. Adding yet another function call into the for loop and hence calling it a couple of million times seems, on the face of it, like it is going to slow things down. Regards, Andy.
On Wed, Apr 01, 2015 at 04:21:50PM +0100, Andy Pont wrote: > Hi David, > > <snipped for brevity> > > > for (i = 0; i < len; i++) { > > /* wait till TX register is empty (TXS == 1) */ > > + start = get_timer(0); > > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & > > OMAP3_MCSPI_CHSTAT_TXS)) { > > - if (--timeout <= 0) { > > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { > > printf("SPI TXS timed out, status=0x%08x\n", > > readl(&ds->regs->channel[ds- > > >slave.cs].chstat)); > > return -1; > > I have a couple of questions... > > Firstly, when in SPL is there access to the get_timer() function? We call timer_init() from board_init_r() in SPL, prior to diving down into loading (or checking for Falcon vs Regular) so this is safe. > Secondly, when using Falcon mode to load Linux directly from SPI (Falcon > mode) then we want to maximise the throughput and save every CPU cycle we > possibly can. Adding yet another function call into the for loop and hence > calling it a couple of million times seems, on the face of it, like it is > going to slow things down. I'd like to see measurements to prove me wrong but this both seems like a bad idea (optimizing by being incorrect, this gives us a correct timeout check like other drivers do) and really unlikely I would think to be noticable. Since we'll be doing the same code-paths in both regular and SPL, trying to time things (by loading a big file) would be easy enough I think. Thanks!
On 7 April 2015 at 05:55, Tom Rini <trini@konsulko.com> wrote: > On Wed, Apr 01, 2015 at 04:21:50PM +0100, Andy Pont wrote: >> Hi David, >> >> <snipped for brevity> >> >> > for (i = 0; i < len; i++) { >> > /* wait till TX register is empty (TXS == 1) */ >> > + start = get_timer(0); >> > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & >> > OMAP3_MCSPI_CHSTAT_TXS)) { >> > - if (--timeout <= 0) { >> > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { >> > printf("SPI TXS timed out, status=0x%08x\n", >> > readl(&ds->regs->channel[ds- >> > >slave.cs].chstat)); >> > return -1; >> >> I have a couple of questions... >> >> Firstly, when in SPL is there access to the get_timer() function? > > We call timer_init() from board_init_r() in SPL, prior to diving down > into loading (or checking for Falcon vs Regular) so this is safe. > >> Secondly, when using Falcon mode to load Linux directly from SPI (Falcon >> mode) then we want to maximise the throughput and save every CPU cycle we >> possibly can. Adding yet another function call into the for loop and hence >> calling it a couple of million times seems, on the face of it, like it is >> going to slow things down. > > I'd like to see measurements to prove me wrong but this both seems like > a bad idea (optimizing by being incorrect, this gives us a correct > timeout check like other drivers do) and really unlikely I would think > to be noticable. Since we'll be doing the same code-paths in both > regular and SPL, trying to time things (by loading a big file) would be > easy enough I think. Thanks! Ping thanks!
As requested: Tested-by: David Dueck <davidcdueck@googlemail.com> Am Freitag, 24. April 2015 schrieb Jagan Teki : > On 7 April 2015 at 05:55, Tom Rini <trini@konsulko.com <javascript:;>> > wrote: > > On Wed, Apr 01, 2015 at 04:21:50PM +0100, Andy Pont wrote: > >> Hi David, > >> > >> <snipped for brevity> > >> > >> > for (i = 0; i < len; i++) { > >> > /* wait till TX register is empty (TXS == 1) */ > >> > + start = get_timer(0); > >> > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & > >> > OMAP3_MCSPI_CHSTAT_TXS)) { > >> > - if (--timeout <= 0) { > >> > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { > >> > printf("SPI TXS timed out, > status=0x%08x\n", > >> > readl(&ds->regs->channel[ds- > >> > >slave.cs].chstat)); > >> > return -1; > >> > >> I have a couple of questions... > >> > >> Firstly, when in SPL is there access to the get_timer() function? > > > > We call timer_init() from board_init_r() in SPL, prior to diving down > > into loading (or checking for Falcon vs Regular) so this is safe. > > > >> Secondly, when using Falcon mode to load Linux directly from SPI (Falcon > >> mode) then we want to maximise the throughput and save every CPU cycle > we > >> possibly can. Adding yet another function call into the for loop and > hence > >> calling it a couple of million times seems, on the face of it, like it > is > >> going to slow things down. > > > > I'd like to see measurements to prove me wrong but this both seems like > > a bad idea (optimizing by being incorrect, this gives us a correct > > timeout check like other drivers do) and really unlikely I would think > > to be noticable. Since we'll be doing the same code-paths in both > > regular and SPL, trying to time things (by loading a big file) would be > > easy enough I think. Thanks! > > Ping > > thanks! > -- > Jagan. >
On 24 April 2015 at 17:04, D. Dueck <davidcdueck@googlemail.com> wrote: > As requested: > Tested-by: David Dueck <davidcdueck@googlemail.com> > > > Am Freitag, 24. April 2015 schrieb Jagan Teki : >> >> On 7 April 2015 at 05:55, Tom Rini <trini@konsulko.com> wrote: >> > On Wed, Apr 01, 2015 at 04:21:50PM +0100, Andy Pont wrote: >> >> Hi David, >> >> >> >> <snipped for brevity> >> >> >> >> > for (i = 0; i < len; i++) { >> >> > /* wait till TX register is empty (TXS == 1) */ >> >> > + start = get_timer(0); >> >> > while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & >> >> > OMAP3_MCSPI_CHSTAT_TXS)) { >> >> > - if (--timeout <= 0) { >> >> > + if (get_timer(start) > SPI_WAIT_TIMEOUT) { >> >> > printf("SPI TXS timed out, >> >> > status=0x%08x\n", >> >> > readl(&ds->regs->channel[ds- >> >> > >slave.cs].chstat)); >> >> > return -1; >> >> >> >> I have a couple of questions... >> >> >> >> Firstly, when in SPL is there access to the get_timer() function? >> > >> > We call timer_init() from board_init_r() in SPL, prior to diving down >> > into loading (or checking for Falcon vs Regular) so this is safe. >> > >> >> Secondly, when using Falcon mode to load Linux directly from SPI >> >> (Falcon >> >> mode) then we want to maximise the throughput and save every CPU cycle >> >> we >> >> possibly can. Adding yet another function call into the for loop and >> >> hence >> >> calling it a couple of million times seems, on the face of it, like it >> >> is >> >> going to slow things down. >> > >> > I'd like to see measurements to prove me wrong but this both seems like >> > a bad idea (optimizing by being incorrect, this gives us a correct >> > timeout check like other drivers do) and really unlikely I would think >> > to be noticable. Since we'll be doing the same code-paths in both >> > regular and SPL, trying to time things (by loading a big file) would be >> > easy enough I think. Thanks! >> >> Ping Applied to u-boot-spi/master thanks!
diff --git a/drivers/spi/omap3_spi.c b/drivers/spi/omap3_spi.c index 651e46e..85f9e85 100644 --- a/drivers/spi/omap3_spi.c +++ b/drivers/spi/omap3_spi.c @@ -20,7 +20,7 @@ #include <asm/io.h> #include "omap3_spi.h" -#define SPI_WAIT_TIMEOUT 3000000 +#define SPI_WAIT_TIMEOUT 10 static void spi_reset(struct omap3_spi_slave *ds) { @@ -227,7 +227,7 @@ int omap3_spi_write(struct spi_slave *slave, unsigned int len, const void *txp, { struct omap3_spi_slave *ds = to_omap3_spi(slave); int i; - int timeout = SPI_WAIT_TIMEOUT; + ulong start; int chconf = readl(&ds->regs->channel[ds->slave.cs].chconf); /* Enable the channel */ @@ -241,9 +241,10 @@ int omap3_spi_write(struct spi_slave *slave, unsigned int len, const void *txp, for (i = 0; i < len; i++) { /* wait till TX register is empty (TXS == 1) */ + start = get_timer(0); while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & OMAP3_MCSPI_CHSTAT_TXS)) { - if (--timeout <= 0) { + if (get_timer(start) > SPI_WAIT_TIMEOUT) { printf("SPI TXS timed out, status=0x%08x\n", readl(&ds->regs->channel[ds->slave.cs].chstat)); return -1; @@ -280,7 +281,7 @@ int omap3_spi_read(struct spi_slave *slave, unsigned int len, void *rxp, { struct omap3_spi_slave *ds = to_omap3_spi(slave); int i; - int timeout = SPI_WAIT_TIMEOUT; + ulong start; int chconf = readl(&ds->regs->channel[ds->slave.cs].chconf); /* Enable the channel */ @@ -295,10 +296,11 @@ int omap3_spi_read(struct spi_slave *slave, unsigned int len, void *rxp, writel(0, &ds->regs->channel[ds->slave.cs].tx); for (i = 0; i < len; i++) { + start = get_timer(0); /* Wait till RX register contains data (RXS == 1) */ while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & OMAP3_MCSPI_CHSTAT_RXS)) { - if (--timeout <= 0) { + if (get_timer(start) > SPI_WAIT_TIMEOUT) { printf("SPI RXS timed out, status=0x%08x\n", readl(&ds->regs->channel[ds->slave.cs].chstat)); return -1; @@ -332,7 +334,7 @@ int omap3_spi_txrx(struct spi_slave *slave, unsigned int len, const void *txp, void *rxp, unsigned long flags) { struct omap3_spi_slave *ds = to_omap3_spi(slave); - int timeout = SPI_WAIT_TIMEOUT; + ulong start; int chconf = readl(&ds->regs->channel[ds->slave.cs].chconf); int irqstatus = readl(&ds->regs->irqstatus); int i=0; @@ -350,9 +352,10 @@ int omap3_spi_txrx(struct spi_slave *slave, unsigned int len, for (i=0; i < len; i++){ /* Write: wait for TX empty (TXS == 1)*/ irqstatus |= (1<< (4*(ds->slave.bus))); + start = get_timer(0); while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & OMAP3_MCSPI_CHSTAT_TXS)) { - if (--timeout <= 0) { + if (get_timer(start) > SPI_WAIT_TIMEOUT) { printf("SPI TXS timed out, status=0x%08x\n", readl(&ds->regs->channel[ds->slave.cs].chstat)); return -1; @@ -368,9 +371,10 @@ int omap3_spi_txrx(struct spi_slave *slave, unsigned int len, writel(((u8 *)txp)[i], tx); /*Read: wait for RX containing data (RXS == 1)*/ + start = get_timer(0); while (!(readl(&ds->regs->channel[ds->slave.cs].chstat) & OMAP3_MCSPI_CHSTAT_RXS)) { - if (--timeout <= 0) { + if (get_timer(start) > SPI_WAIT_TIMEOUT) { printf("SPI RXS timed out, status=0x%08x\n", readl(&ds->regs->channel[ds->slave.cs].chstat)); return -1;
The timeout value is never reset during the transfer. This means that when transferring more data we eventually trigger the timeout. This was reported on the mailing list: "Spansion SPI flash read timeout with AM335x" Signed-off-by: David Dueck <davidcdueck@googlemail.com> CC: Tom Rini <trini@konsulko.com> CC: Jagannadh Teki <jagannadh.teki@gmail.com> CC: Stefan Roese <sr@denx.de> CC: Andy Pont <andy.pont@sdcsystems.com> --- Changes since v1: - fix style issue - fix CC line drivers/spi/omap3_spi.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-)