| Message ID | CWXP265MB2680A8DD0FFA6FECBFDFB027C4579@CWXP265MB2680.GBRP265.PROD.OUTLOOK.COM |
|---|---|
| State | New |
| Headers | show |
| Series | [v2] mmc: block: ioctl: No busywaiting of non-TRAN CMDs | expand |
On Fri, 7 May 2021 at 12:34, Christian Löhle <CLoehle@hyperstone.com> wrote: > > Prevent busywaiting for TRAN state indication > after issuing a command that will not transition > to TRAN state. > > Signed-off-by: Christian Loehle <cloehle@hyperstone.com> > --- > drivers/mmc/core/block.c | 17 ++++++++++++++++- > 1 file changed, 16 insertions(+), 1 deletion(-) > > diff --git a/drivers/mmc/core/block.c b/drivers/mmc/core/block.c > index 689eb9afeeed..48be2ca5e3d1 100644 > --- a/drivers/mmc/core/block.c > +++ b/drivers/mmc/core/block.c > @@ -446,6 +446,20 @@ static int card_busy_detect(struct mmc_card *card, unsigned int timeout_ms, > return err; > } > > +static inline bool is_tran_transition_cmd(struct mmc_command *cmd, > + struct mmc_card *card) > +{ > + /* Cards will not be in TRAN after completing identification commands > + * or MMC_SEND_STATUS if they are not selected. > + */ > + return !(cmd->opcode == MMC_SEND_CID > + || cmd->opcode == MMC_ALL_SEND_CID > + || cmd->opcode == MMC_SEND_CSD > + || (cmd->opcode == MMC_SEND_STATUS && > + MMC_EXTRACT_INDEX_FROM_ARG(cmd->arg) != card->rca)); > + > +} > + > static int __mmc_blk_ioctl_cmd(struct mmc_card *card, struct mmc_blk_data *md, > struct mmc_blk_ioc_data *idata) > { > @@ -593,7 +607,8 @@ static int __mmc_blk_ioctl_cmd(struct mmc_card *card, struct mmc_blk_data *md, > > memcpy(&(idata->ic.response), cmd.resp, sizeof(cmd.resp)); > > - if (idata->rpmb || (cmd.flags & MMC_RSP_R1B) == MMC_RSP_R1B) { > + if ((idata->rpmb || (cmd.flags & MMC_RSP_R1B)) > + && is_tran_transition_cmd(&cmd, card)) { None of the commands you are checking for should have an R1B response according to the spec, I think. That said, I don't think we should do these kinds of sanity checks in the kernel for the mmc ioctls, that just doesn't scale. > /* > * Ensure RPMB/R1B command has completed by polling CMD13 > * "Send Status". > -- Kind regards Uffe
On Friday, May 7, 2021 1:34 PM,Ulf Hansson <ulf.hansson@linaro.org> wrote: > > None of the commands you are checking for should have an R1B response > according to the spec, I think. > > That said, I don't think we should do these kinds of sanity checks in > the kernel for the mmc ioctls, that just doesn't scale. You are absolutely correct, my bad, I had a userspace program setting the flags wrong. Even for a SEND_STATUS R1B is only expected if the card was not selected and should be set accordingly by the userspace. Regards, Christian Hyperstone GmbH | Line-Eid-Strasse 3 | 78467 Konstanz Managing Directors: Dr. Jan Peter Berns. Commercial register of local courts: Freiburg HRB381782
diff --git a/drivers/mmc/core/block.c b/drivers/mmc/core/block.c index 689eb9afeeed..48be2ca5e3d1 100644 --- a/drivers/mmc/core/block.c +++ b/drivers/mmc/core/block.c @@ -446,6 +446,20 @@ static int card_busy_detect(struct mmc_card *card, unsigned int timeout_ms, return err; } +static inline bool is_tran_transition_cmd(struct mmc_command *cmd, + struct mmc_card *card) +{ + /* Cards will not be in TRAN after completing identification commands + * or MMC_SEND_STATUS if they are not selected. + */ + return !(cmd->opcode == MMC_SEND_CID + || cmd->opcode == MMC_ALL_SEND_CID + || cmd->opcode == MMC_SEND_CSD + || (cmd->opcode == MMC_SEND_STATUS && + MMC_EXTRACT_INDEX_FROM_ARG(cmd->arg) != card->rca)); + +} + static int __mmc_blk_ioctl_cmd(struct mmc_card *card, struct mmc_blk_data *md, struct mmc_blk_ioc_data *idata) { @@ -593,7 +607,8 @@ static int __mmc_blk_ioctl_cmd(struct mmc_card *card, struct mmc_blk_data *md, memcpy(&(idata->ic.response), cmd.resp, sizeof(cmd.resp)); - if (idata->rpmb || (cmd.flags & MMC_RSP_R1B) == MMC_RSP_R1B) { + if ((idata->rpmb || (cmd.flags & MMC_RSP_R1B)) + && is_tran_transition_cmd(&cmd, card)) { /* * Ensure RPMB/R1B command has completed by polling CMD13 * "Send Status".
Prevent busywaiting for TRAN state indication after issuing a command that will not transition to TRAN state. Signed-off-by: Christian Loehle <cloehle@hyperstone.com> --- drivers/mmc/core/block.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)