| Message ID | 20170530090346.19088-1-wsa+renesas@sang-engineering.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Tue, May 30, 2017 at 11:03:46AM +0200, Wolfram Sang wrote: > Most registers need to wait until the command is completed, not > necessarily until the bus is free. At least, R-Car 2+ SoCs can signal > that via the CBSY bit, so let's use it there instead of SCLKDIVEN to > save a little bit of delay. > > Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> > --- > > Tested with a Renesas Lager board (H2). Moved big files around across different > SD cards, re- inserted cards multiple times. All worked fine. > > @Dirk: can you do some additional testing for your use case? That would be much > appreciated! OK, since Dirk told me he'll need some more time to test it anyhow and Simon meanwhile sent some cleanup patches, I'll simply post an updated version of this patch once Simon's cleanup patches hit mmc/next. I think this will be easiest.
On Tue, May 30, 2017 at 10:02:18PM +0200, Wolfram Sang wrote: > On Tue, May 30, 2017 at 11:03:46AM +0200, Wolfram Sang wrote: > > Most registers need to wait until the command is completed, not > > necessarily until the bus is free. At least, R-Car 2+ SoCs can signal > > that via the CBSY bit, so let's use it there instead of SCLKDIVEN to > > save a little bit of delay. > > > > Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> > > --- > > > > Tested with a Renesas Lager board (H2). Moved big files around across different > > SD cards, re- inserted cards multiple times. All worked fine. > > > > @Dirk: can you do some additional testing for your use case? That would be much > > appreciated! > > OK, since Dirk told me he'll need some more time to test it anyhow and > Simon meanwhile sent some cleanup patches, I'll simply post an updated > version of this patch once Simon's cleanup patches hit mmc/next. I think > this will be easiest. Thanks, sorry for the clash - we seem to often hit the SDHI driver at about the same time. -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> Thanks, sorry for the clash - we seem to often hit the SDHI driver > at about the same time. Yes, but given the amount of refactoring done, the rebase work is surprisingly small. So, I am all fine :) Thanks!
diff --git a/drivers/mmc/host/renesas_sdhi_core.c b/drivers/mmc/host/renesas_sdhi_core.c index 82150a96639195..288153b5f6835e 100644 --- a/drivers/mmc/host/renesas_sdhi_core.c +++ b/drivers/mmc/host/renesas_sdhi_core.c @@ -395,12 +395,14 @@ static void renesas_sdhi_hw_reset(struct tmio_mmc_host *host) sd_scc_read32(host, priv, SH_MOBILE_SDHI_SCC_RVSCNTL)); } -static int renesas_sdhi_wait_idle(struct tmio_mmc_host *host) +static int renesas_sdhi_wait_idle(struct tmio_mmc_host *host, u32 bit) { int timeout = 1000; + /* CBSY is set when busy, SCLKDIVEN is cleared when busy */ + u32 wait_state = (bit == TMIO_STAT_CMD_BUSY ? TMIO_STAT_CMD_BUSY : 0); - while (--timeout && !(sd_ctrl_read16_and_16_as_32(host, CTL_STATUS) - & TMIO_STAT_SCLKDIVEN)) + while (--timeout && (sd_ctrl_read16_and_16_as_32(host, CTL_STATUS) + & bit) == wait_state) udelay(1); if (!timeout) { @@ -413,18 +415,23 @@ static int renesas_sdhi_wait_idle(struct tmio_mmc_host *host) static int renesas_sdhi_write16_hook(struct tmio_mmc_host *host, int addr) { + u32 bit = TMIO_STAT_SCLKDIVEN; + switch (addr) { case CTL_SD_CMD: case CTL_STOP_INTERNAL_ACTION: case CTL_XFER_BLK_COUNT: - case CTL_SD_CARD_CLK_CTL: case CTL_SD_XFER_LEN: case CTL_SD_MEM_CARD_OPT: case CTL_TRANSACTION_CTL: case CTL_DMA_ENABLE: case EXT_ACC: - return renesas_sdhi_wait_idle(host); + if (host->pdata->flags & TMIO_MMC_MIN_RCAR2) + bit = TMIO_STAT_CMD_BUSY; + /* fallthrough */ + case CTL_SD_CARD_CLK_CTL: + return renesas_sdhi_wait_idle(host, bit); } return 0;
Most registers need to wait until the command is completed, not necessarily until the bus is free. At least, R-Car 2+ SoCs can signal that via the CBSY bit, so let's use it there instead of SCLKDIVEN to save a little bit of delay. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> --- Tested with a Renesas Lager board (H2). Moved big files around across different SD cards, re- inserted cards multiple times. All worked fine. @Dirk: can you do some additional testing for your use case? That would be much appreciated! Changes since RFC (from Jan 2016): * rebased to latest code base * use existing bit defines instead of creating new ones drivers/mmc/host/renesas_sdhi_core.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-)