Messages in this thread | | | Subject | Re: [PATCH] mtd: dc21285.c: handle nw_gpio_lock correctly | From | Artem Bityutskiy <> | Date | Sat, 26 May 2012 16:54:24 +0300 |
| |
On Fri, 2012-05-25 at 10:28 +0200, Christian Dietrich wrote: > nw_gpio_lock is a raw_spinlock_t, therefore raw_spin_lock_irqsave should be > used. it doesn't make a difference, while rlock is the first element of > spinlock_t. > > Also the check for machine_is_netwinder() is a double check of > CONFIG_ARCH_NETWINDER. > > Signed-off-by: Christian Dietrich <christian.dietrich@informatik.uni-erlangen.de>
I do not understand this commit message. Please, split your patch on 2 patches and put a better to each one. Thanks!
> --- > drivers/mtd/maps/dc21285.c | 13 +++++-------- > 1 files changed, 5 insertions(+), 8 deletions(-) > > diff --git a/drivers/mtd/maps/dc21285.c b/drivers/mtd/maps/dc21285.c > index 080f060..38fbf23 100644 > --- a/drivers/mtd/maps/dc21285.c > +++ b/drivers/mtd/maps/dc21285.c > @@ -38,9 +38,9 @@ static void nw_en_write(void) > * we want to write a bit pattern XXX1 to Xilinx to enable > * the write gate, which will be open for about the next 2ms. > */ > - spin_lock_irqsave(&nw_gpio_lock, flags); > + raw_spin_lock_irqsave(&nw_gpio_lock, flags);
Please, make this to be a separate patch.
And surely there are many other places in the kernel which need this conversion?
> static void dc21285_write8(struct map_info *map, const map_word d, unsigned long adr) > { > - if (machine_is_netwinder()) > - nw_en_write(); > + nw_en_write(); > *CSR_ROMWRITEREG = adr & 3;
I do not understand why this "if" statement is not needed? Why it was there in the first place and why you remove it? Please, describe this in the commit message.
-- Best Regards, Artem Bityutskiy [unhandled content-type:application/pgp-signature] | |