| Message ID | 20231129161459.1002323-17-andriy.shevchenko@linux.intel.com |
|---|---|
| State | Accepted |
| Commit | 390270f25b414fd54b307cd68851b36b52f952b5 |
| Headers | show |
| Series | pinctrl: Convert struct group_desc to use struct pingroup | expand |
Hi Andy, On Wed, Nov 29, 2023 at 1:15 PM Andy Shevchenko <andriy.shevchenko@linux.intel.com> wrote: > > Convert drivers to use grp member embedded in struct group_desc. You explained what you did, but you did not explain the reason.
On Wed, Nov 29, 2023 at 01:36:49PM -0300, Fabio Estevam wrote: > Hi Andy, > > On Wed, Nov 29, 2023 at 1:15 PM Andy Shevchenko > <andriy.shevchenko@linux.intel.com> wrote: > > > > Convert drivers to use grp member embedded in struct group_desc. > > You explained what you did, but you did not explain the reason. It's explained in the first paragraph in the cover letter. Do you want to copy this into each commit message? Note, `b4 shazam ...` has an ability to consider patch series as Git PR, where the cover letter goes as template for the merge commit message. That's how I plan to merge it to my tree.
On Wed, Nov 29, 2023 at 02:08:38PM -0300, Fabio Estevam wrote: > On Wed, Nov 29, 2023 at 2:01 PM Andy Shevchenko > <andriy.shevchenko@linux.intel.com> wrote: > > > > It's explained in the first paragraph in the cover letter. Do you > > want to copy this into each commit message? > > Yes, much better to have the information into each commit message. Here it would be like "Because other members will be removed to avoid duplication and desynchronisation of the generic pin group description." Linus, what do you think about this?
On Wed, Nov 29, 2023 at 10:41:14PM +0100, Linus Walleij wrote: > On Wed, Nov 29, 2023 at 6:18 PM Andy Shevchenko > <andriy.shevchenko@linux.intel.com> wrote: > > On Wed, Nov 29, 2023 at 02:08:38PM -0300, Fabio Estevam wrote: > > > On Wed, Nov 29, 2023 at 2:01 PM Andy Shevchenko > > > <andriy.shevchenko@linux.intel.com> wrote: > > > > > > > > It's explained in the first paragraph in the cover letter. Do you > > > > want to copy this into each commit message? > > > > > > Yes, much better to have the information into each commit message. > > > > Here it would be like > > "Because other members will be removed to avoid duplication and > > desynchronisation of the generic pin group description." > > > > Linus, what do you think about this? > > I can just add that to each commit while applying if it makes everyone happy. I'm pretty much fine, but I dunno if you are using `b4 shazam -H ...` I found that very cool feature. > No need to resend for that. Yeah, we may utilise `git msg-filter ...` to fulfill the job (for curious one, it can be run even if we are in the middle of `git rebase --interactive ...`).
diff --git a/drivers/pinctrl/freescale/pinctrl-imx.c b/drivers/pinctrl/freescale/pinctrl-imx.c index 28164b04b1be..2d3d80921c0d 100644 --- a/drivers/pinctrl/freescale/pinctrl-imx.c +++ b/drivers/pinctrl/freescale/pinctrl-imx.c @@ -42,7 +42,7 @@ static inline const struct group_desc *imx_pinctrl_find_group_by_name( for (i = 0; i < pctldev->num_groups; i++) { grp = pinctrl_generic_get_group(pctldev, i); - if (grp && !strcmp(grp->name, name)) + if (grp && !strcmp(grp->grp.name, name)) break; } @@ -79,9 +79,9 @@ static int imx_dt_node_to_map(struct pinctrl_dev *pctldev, } if (info->flags & IMX_USE_SCU) { - map_num += grp->num_pins; + map_num += grp->grp.npins; } else { - for (i = 0; i < grp->num_pins; i++) { + for (i = 0; i < grp->grp.npins; i++) { pin = &((struct imx_pin *)(grp->data))[i]; if (!(pin->conf.mmio.config & IMX_NO_PAD_CTL)) map_num++; @@ -109,7 +109,7 @@ static int imx_dt_node_to_map(struct pinctrl_dev *pctldev, /* create config map */ new_map++; - for (i = j = 0; i < grp->num_pins; i++) { + for (i = j = 0; i < grp->grp.npins; i++) { pin = &((struct imx_pin *)(grp->data))[i]; /* @@ -263,10 +263,10 @@ static int imx_pmx_set(struct pinctrl_dev *pctldev, unsigned selector, if (!func) return -EINVAL; - npins = grp->num_pins; + npins = grp->grp.npins; dev_dbg(ipctl->dev, "enable function %s group %s\n", - func->name, grp->name); + func->name, grp->grp.name); for (i = 0; i < npins; i++) { /* @@ -423,7 +423,7 @@ static void imx_pinconf_group_dbg_show(struct pinctrl_dev *pctldev, if (!grp) return; - for (i = 0; i < grp->num_pins; i++) { + for (i = 0; i < grp->grp.npins; i++) { struct imx_pin *pin = &((struct imx_pin *)(grp->data))[i]; name = pin_get_name(pctldev, pin->pin); @@ -526,7 +526,7 @@ static int imx_pinctrl_parse_groups(struct device_node *np, pin_size = FSL_PIN_SIZE; /* Initialise group */ - grp->name = np->name; + grp->grp.name = np->name; /* * the binding format is fsl,pins = <PIN_FUNC_ID CONFIG ...>, @@ -554,19 +554,17 @@ static int imx_pinctrl_parse_groups(struct device_node *np, return -EINVAL; } - grp->num_pins = size / pin_size; - grp->data = devm_kcalloc(ipctl->dev, - grp->num_pins, sizeof(struct imx_pin), - GFP_KERNEL); + grp->grp.npins = size / pin_size; + grp->data = devm_kcalloc(ipctl->dev, grp->grp.npins, sizeof(*pin), GFP_KERNEL); if (!grp->data) return -ENOMEM; - pins = devm_kcalloc(ipctl->dev, grp->num_pins, sizeof(*pins), GFP_KERNEL); + pins = devm_kcalloc(ipctl->dev, grp->grp.npins, sizeof(*pins), GFP_KERNEL); if (!pins) return -ENOMEM; - grp->pins = pins; + grp->grp.pins = pins; - for (i = 0; i < grp->num_pins; i++) { + for (i = 0; i < grp->grp.npins; i++) { pin = &((struct imx_pin *)(grp->data))[i]; if (info->flags & IMX_USE_SCU) info->imx_pinctrl_parse_pin(ipctl, &pins[i], pin, &list); @@ -613,8 +611,7 @@ static int imx_pinctrl_parse_functions(struct device_node *np, i = 0; for_each_child_of_node(np, child) { - grp = devm_kzalloc(ipctl->dev, sizeof(struct group_desc), - GFP_KERNEL); + grp = devm_kzalloc(ipctl->dev, sizeof(*grp), GFP_KERNEL); if (!grp) { of_node_put(child); return -ENOMEM;
Convert drivers to use grp member embedded in struct group_desc. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> --- drivers/pinctrl/freescale/pinctrl-imx.c | 31 +++++++++++-------------- 1 file changed, 14 insertions(+), 17 deletions(-)