[patch] ath9k_hw: min_t() casts u32 to int

[patch] ath9k_hw: min_t() casts u32 to int

[Dan Carpenter]

The code here treats very large values of "limit" as less than
MAX_POWER_RATE because of the cast to int.  We should do the compare
as u32 instead.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>

diff --git a/drivers/net/wireless/ath/ath9k/hw.c b/drivers/net/wireless/ath/ath9k/hw.c
index e0c8549..f9abbb7 100644
--- a/drivers/net/wireless/ath/ath9k/hw.c
+++ b/drivers/net/wireless/ath/ath9k/hw.c
@@ -2526,7 +2526,7 @@ void ath9k_hw_set_txpowerlimit(struct ath_hw *ah, u32 limit, bool test)
 	struct ath9k_channel *chan = ah->curchan;
 	struct ieee80211_channel *channel = chan->chan;
 
-	reg->power_limit = min_t(int, limit, MAX_RATE_POWER);
+	reg->power_limit = min_t(u32, limit, MAX_RATE_POWER);
 	if (test)
 		channel->max_power = MAX_RATE_POWER / 2;
 

Re: [patch] ath9k_hw: min_t() casts u32 to int

[Pavel Roskin]

On Mon, 17 Oct 2011 10:28:23 +0300
Dan Carpenter <dan.carpenter@oracle.com> wrote:

> The code here treats very large values of "limit" as less than
> MAX_POWER_RATE because of the cast to int.  We should do the compare
> as u32 instead.
> 
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>

Reviewed-by: Pavel Roskin <proski@gnu.org>

I don't think it can actually happen.  ath9k_hw_set_txpowerlimit() is
called twice.  In one case, limit is MAX_RATE_POWER, in another case,
it's an argument of type u16.  Still, it's better not to have the code
that looks wrong.

ath9k_cmn_update_txpow() also has code that _looks_ wrong.  *txpower is
not assigned a value if the new power happens to be equal to the
current one.  Also, the second argument (cur_txpow) is not used.
Returning a value by a pointer seems unnecessary if the function returns
void.  It could simply return the new tx power.  I'll submit a patch
shortly.

-- 
Regards,
Pavel Roskin