From: Bamvor Jian Zhang <bamvor.zhangjian@linaro.org>
To: julien.grossholtz@savoirfairelinux.com
Cc: linux-gpio@vger.kernel.org, linus.walleij@linaro.org,
broonie@kernel.org, kernel@savoirfairelinux.com, arnd@arndb.de,
bamvor.zhangjian@linaro.org
Subject: Re: [Kernel] [PATCH] gpiolib: rewrite gpiochip_add_to_list
Date: Mon, 11 Jan 2016 22:35:13 +0800 [thread overview]
Message-ID: <1452522913-6298-1-git-send-email-bamvor.zhangjian@linaro.org> (raw)
In-Reply-To: <CAFy1USRc8=wwpr9rGkcvYR5gdO-cz1c1onfngSAECb94XpmGgg@mail.gmail.com>
Hi, Julien
On 01/09/2016 11:16 PM, Bamvor Zhang Jian wrote:
> Hi, Julien
>
> On 01/09/2016 06:56 AM, Julien Grossholtz wrote:
>> Hi,
>>
>> The issue is in the break condition of the list_for_each_entry_safe loop.
>> If there is a chip at the beginning it breaks even if there is another
>> one after. You should check the next chip:
>> if (chip->base > prev->base && chip->base + chip->ngpio <= next->base)
> Thanks you suggestion. It indeed fix the issue you mentioned.
> I will send the new version after I make sure there is no other corner case.
Could you do me favor to test the following patches?
It works for the following ranges: [0,31], [32,64], [-1,32], [-1, 32] and the
other test cases I wrote.
Regards
Bamvor
---
drivers/gpio/gpiolib.c | 65 ++++++++++++++++++++++----------------------------
1 file changed, 28 insertions(+), 37 deletions(-)
diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 3db34e7..9460f94 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -189,55 +189,46 @@ EXPORT_SYMBOL_GPL(gpiod_get_direction);
*/
static int gpiochip_add_to_list(struct gpio_chip *chip)
{
- struct gpio_chip *iterator;
- struct gpio_chip *previous = NULL;
+ struct gpio_chip *prev, *next;
if (list_empty(&gpio_chips)) {
- list_add_tail(&chip->list, &gpio_chips);
+ /* initial entry in list */
+ list_add(&chip->list, &gpio_chips);
return 0;
}
- list_for_each_entry(iterator, &gpio_chips, list) {
- if (iterator->base >= chip->base + chip->ngpio) {
- /*
- * Iterator is the first GPIO chip so there is no
- * previous one
- */
- if (!previous) {
- goto found;
- } else {
- /*
- * We found a valid range(means
- * [base, base + ngpio - 1]) between previous
- * and iterator chip.
- */
- if (previous->base + previous->ngpio
- <= chip->base)
- goto found;
- }
- }
- previous = iterator;
+ next = list_entry(gpio_chips.next, struct gpio_chip, list);
+ if (chip->base + chip->ngpio <= next->base) {
+ /* add before first entry */
+ list_add(&chip->list, &gpio_chips);
+ return 0;
}
- /*
- * We are beyond the last chip in the list and iterator now
- * points to the head.
- * Let iterator point to the last chip in the list.
- */
+ prev = list_entry(gpio_chips.prev, struct gpio_chip, list);
+ if (prev->base + prev->ngpio <= chip->base) {
+ /* add behind last entry */
+ list_add_tail(&chip->list, &gpio_chips);
+ return 0;
+ }
- iterator = list_last_entry(&gpio_chips, struct gpio_chip, list);
- if (iterator->base + iterator->ngpio <= chip->base)
- goto found;
+ list_for_each_entry_safe(prev, next, &gpio_chips, list) {
+ /* at the end of the list */
+ if (&(next->list) == &gpio_chips)
+ break;
- dev_err(chip->parent,
- "GPIO integer space overlap, cannot add chip\n");
- return -EBUSY;
+ /* add between prev and next */
+ if ( prev->base + prev->ngpio <= chip->base
+ && chip->base + chip->ngpio <= next->base) {
+ list_add(&chip->list, &prev->list);
+ return 0;
+ }
+ }
-found:
- list_add_tail(&chip->list, &iterator->list);
- return 0;
+ dev_err(chip->parent, "GPIO integer space overlap, cannot add chip\n");
+ return -EBUSY;
}
+
/**
* Convert a GPIO name to its descriptor
*/
--
2.1.4
next prev parent reply other threads:[~2016-01-11 14:36 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-01-08 6:37 [PATCH] gpiolib: rewrite gpiochip_add_to_list Bamvor Jian Zhang
2016-01-08 6:37 ` Bamvor Jian Zhang
2016-01-08 21:47 ` Julien Grossholtz
2016-01-08 22:56 ` [Kernel] " Julien Grossholtz
2016-01-09 15:16 ` Bamvor Zhang Jian
2016-01-11 14:35 ` Bamvor Jian Zhang [this message]
2016-01-11 19:56 ` Julien Grossholtz
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1452522913-6298-1-git-send-email-bamvor.zhangjian@linaro.org \
--to=bamvor.zhangjian@linaro.org \
--cc=arnd@arndb.de \
--cc=broonie@kernel.org \
--cc=julien.grossholtz@savoirfairelinux.com \
--cc=kernel@savoirfairelinux.com \
--cc=linus.walleij@linaro.org \
--cc=linux-gpio@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).