Kees Cook <keesc...@chromium.org> writes:
> On Wed, Jul 31, 2019 at 12:28:55AM +1000, Michael Ellerman wrote:
>> Stephen Rothwell <s...@canb.auug.org.au> writes:
>> > Mark switch cases where we are expecting to fall through.
>> >
>> > This patch fixes the following warning (Building: powerpc):
>> >
>> > drivers/macintosh/smu.c: In function 'smu_queue_i2c':
>> > drivers/macintosh/smu.c:854:21: warning: this statement may fall through
>> > [-Wimplicit-fallthrough=]
>> > cmd->info.devaddr &= 0xfe;
>> > ~~~~~~~~~~~~~~~~~~^~~~~~~
>> > drivers/macintosh/smu.c:855:2: note: here
>> > case SMU_I2C_TRANSFER_STDSUB:
>> > ^~~~
>> >
>> > Cc: Benjamin Herrenschmidt <b...@kernel.crashing.org>
>> > Cc: Gustavo A. R. Silva <gust...@embeddedor.com>
>> > Cc: Kees Cook <keesc...@chromium.org>
>> > Signed-off-by: Stephen Rothwell <s...@canb.auug.org.au>
>> > ---
>> > drivers/macintosh/smu.c | 1 +
>> > 1 file changed, 1 insertion(+)
>> >
>> > diff --git a/drivers/macintosh/smu.c b/drivers/macintosh/smu.c
>> > index 276065c888bc..23f1f41c8602 100644
>> > --- a/drivers/macintosh/smu.c
>> > +++ b/drivers/macintosh/smu.c
>> > @@ -852,6 +852,7 @@ int smu_queue_i2c(struct smu_i2c_cmd *cmd)
>> > break;
>> > case SMU_I2C_TRANSFER_COMBINED:
>> > cmd->info.devaddr &= 0xfe;
>> > + /* fall through */
>> > case SMU_I2C_TRANSFER_STDSUB:
>> > if (cmd->info.sublen > 3)
>> > return -EINVAL;
>>
>> Why do we think it's an expected fall through? I can't really convince
>> myself from the surrounding code that it's definitely intentional.
>
> Yeah, good question. Just now when I went looking for who
> used SMU_I2C_TRANSFER_COMBINED, I found the only caller in
> arch/powerpc/platforms/powermac/low_i2c.c and it is clearly using a
> fall-through for building the command for "stdsub" and "combined",
> so I think that's justification enough:
>
> switch(bus->mode) {
> case pmac_i2c_mode_std:
> if (subsize != 0)
> return -EINVAL;
> cmd->info.type = SMU_I2C_TRANSFER_SIMPLE;
> break;
> case pmac_i2c_mode_stdsub:
> case pmac_i2c_mode_combined:
> if (subsize > 3 || subsize < 1)
> return -EINVAL;
> cmd->info.sublen = subsize;
> /* that's big-endian only but heh ! */
> memcpy(&cmd->info.subaddr, ((char *)&subaddr) + (4 - subsize),
> subsize);
> if (bus->mode == pmac_i2c_mode_stdsub)
> cmd->info.type = SMU_I2C_TRANSFER_STDSUB;
> else
> cmd->info.type = SMU_I2C_TRANSFER_COMBINED;
>
>
> Reviewed-by: Kees Cook <keesc...@chromium.org>
Thanks.
cheers
