| Message ID | 20190620152432.1408278-1-lkundrak@v3.sk (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [RESEND] mmc: core: try to use an id from the devicetree | expand |
+ Doug On Thu, 20 Jun 2019 at 17:24, Lubomir Rintel <lkundrak@v3.sk> wrote: > > If there's a mmc* alias in the device tree, take the device number from > it, so that we end up with a device name that matches the alias. Lots of people would be happy if I queue something along the lines of what you propose. I am not really having any big problems with it, but I am reluctant to queue it because of other peoples quite strong opinions [1] that have been expressed in regards to this already. Kind regards Uffe [1] https://www.spinics.net/lists/devicetree-spec/msg00254.html > > Signed-off-by: Lubomir Rintel <lkundrak@v3.sk> > --- > drivers/mmc/core/host.c | 7 ++++++- > 1 file changed, 6 insertions(+), 1 deletion(-) > > diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c > index 6a51f7a06ce7..4733ddb894da 100644 > --- a/drivers/mmc/core/host.c > +++ b/drivers/mmc/core/host.c > @@ -411,7 +411,12 @@ struct mmc_host *mmc_alloc_host(int extra, struct device *dev) > /* scanning will be enabled when we're ready */ > host->rescan_disable = 1; > > - err = ida_simple_get(&mmc_host_ida, 0, 0, GFP_KERNEL); > + /* prefer an alias */ > + err = of_alias_get_id(dev->of_node, "mmc"); > + if (err < 0) > + err = 0; > + > + err = ida_simple_get(&mmc_host_ida, err, 0, GFP_KERNEL); > if (err < 0) { > kfree(host); > return NULL; > -- > 2.21.0 >
Hi, On Thu, Jun 20, 2019 at 8:37 AM Ulf Hansson <ulf.hansson@linaro.org> wrote: > > + Doug > > On Thu, 20 Jun 2019 at 17:24, Lubomir Rintel <lkundrak@v3.sk> wrote: > > > > If there's a mmc* alias in the device tree, take the device number from > > it, so that we end up with a device name that matches the alias. > > Lots of people would be happy if I queue something along the lines of > what you propose. I am not really having any big problems with it, but > I am reluctant to queue it because of other peoples quite strong > opinions [1] that have been expressed in regards to this already. > > Kind regards > Uffe > > [1] > https://www.spinics.net/lists/devicetree-spec/msg00254.html Yeah, I personally like being able to assign numbers too, but unfortunately there are lots of people who objected. BTW: if you prefer the patchwork view of the same discussion that Ulf pointed to: https://lore.kernel.org/patchwork/cover/674381/ As per that discussion, I think might be OK if we could find a way to assign a string-based name to devices. Then if your user manual calls them "emmc", "sd", and "sdio" you could name them that way. ...and if your manual calls them "emmc", "sd0", "sd1" you could name them that way. ...but I wouldn't swear that people would actually truly like that. Given the total number of people who keep feeling like this is an issue that needs to be solved, though, it does seem worthwhile for someone to come up with a solution. -Doug
diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c index 6a51f7a06ce7..4733ddb894da 100644 --- a/drivers/mmc/core/host.c +++ b/drivers/mmc/core/host.c @@ -411,7 +411,12 @@ struct mmc_host *mmc_alloc_host(int extra, struct device *dev) /* scanning will be enabled when we're ready */ host->rescan_disable = 1; - err = ida_simple_get(&mmc_host_ida, 0, 0, GFP_KERNEL); + /* prefer an alias */ + err = of_alias_get_id(dev->of_node, "mmc"); + if (err < 0) + err = 0; + + err = ida_simple_get(&mmc_host_ida, err, 0, GFP_KERNEL); if (err < 0) { kfree(host); return NULL;
If there's a mmc* alias in the device tree, take the device number from it, so that we end up with a device name that matches the alias. Signed-off-by: Lubomir Rintel <lkundrak@v3.sk> --- drivers/mmc/core/host.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)