netdev - Re: [PATCH net-next] net: ipa: always inline ipa_aggr_granularity

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 11 Aug 2021 07:06:01 -0500
From:   Alex Elder <elder@...aro.org>
To:     Leon Romanovsky <leon@...nel.org>
Cc:     davem@...emloft.net, kuba@...nel.org, bjorn.andersson@...aro.org,
        evgreen@...omium.org, cpratapa@...eaurora.org,
        subashab@...eaurora.org, lkp@...el.com, elder@...nel.org,
        netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH net-next] net: ipa: always inline
 ipa_aggr_granularity_val()

On 8/11/21 7:02 AM, Leon Romanovsky wrote:
> On Tue, Aug 10, 2021 at 11:02:13AM -0500, Alex Elder wrote:
>> It isn't required, but all callers of ipa_aggr_granularity_val()
>> pass a constant value (IPA_AGGR_GRANULARITY) as the usec argument.
>> Two of those callers are in ipa_validate_build(), with the result
>> being passed to BUILD_BUG_ON().
>>
>> Evidently the "sparc64-linux-gcc" compiler (at least) doesn't always
>> inline ipa_aggr_granularity_val(), so the result of the function is
>> not constant at compile time, and that leads to build errors.
>>
>> Define the function with the __always_inline attribute to avoid the
>> errors.  And given that the function is inline, we can switch the
>> WARN_ON() there to be BUILD_BUG_ON().
>>
>> Fixes: 5bc5588466a1f ("net: ipa: use WARN_ON() rather than assertions")
>> Reported-by: kernel test robot <lkp@...el.com>
>> Signed-off-by: Alex Elder <elder@...aro.org>
>> ---
>>
>> David/Jakub, this fixes a bug in a commit in net-next/master.  -Alex
>>
>>  drivers/net/ipa/ipa_main.c | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/net/ipa/ipa_main.c b/drivers/net/ipa/ipa_main.c
>> index 25bbb456e0078..f90b3521e266b 100644
>> --- a/drivers/net/ipa/ipa_main.c
>> +++ b/drivers/net/ipa/ipa_main.c
>> @@ -255,9 +255,9 @@ ipa_hardware_config_qsb(struct ipa *ipa, const struct ipa_data *data)
>>   * less than the number of timer ticks in the requested period.  0 is not
>>   * a valid granularity value.
>>   */
>> -static u32 ipa_aggr_granularity_val(u32 usec)
>> +static __always_inline u32 ipa_aggr_granularity_val(u32 usec)
>>  {
>> -	WARN_ON(!usec);
>> +	BUILD_BUG_ON(!usec);
> 
> So what exactly are you checking here if all callers pass same value?
> It is in-kernel API, declared as static inside one module. There is no
> need to protect from itself.

Yeah that's a good point.  It can just as well be removed.
I think the check was added before I knew it was only going
to be used with a single constant value.  That said, the
point was to check at runtime a required constraint.

I'll post version 2 that simply removes it.  Thanks.

					-Alex

> 
> Thanks
> 
>>  
>>  	return DIV_ROUND_CLOSEST(usec * TIMER_FREQUENCY, USEC_PER_SEC) - 1;
>>  }
>> -- 
>> 2.27.0
>>