Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <0ba56650-e1bf-5a8f-86cd-8406f3dc75be@linaro.org>
Date:   Wed, 2 Nov 2022 07:44:35 -0500
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.2.2
Subject: Re: [PATCH net-next 9/9] net: ipa: use a bitmap for enabled endpoints
Content-Language: en-US
To:     Jakub Kicinski <kuba@kernel.org>
Cc:     davem@davemloft.net, edumazet@google.com, pabeni@redhat.com,
        mka@chromium.org, evgreen@chromium.org, andersson@kernel.org,
        quic_cpratapa@quicinc.com, quic_avuyyuru@quicinc.com,
        quic_jponduru@quicinc.com, quic_subashab@quicinc.com,
        elder@kernel.org, netdev@vger.kernel.org,
        linux-arm-msm@vger.kernel.org, linux-kernel@vger.kernel.org
References: <20221030001828.754010-1-elder@linaro.org>
 <20221030001828.754010-10-elder@linaro.org>
 <20221101213404.3e4c3b8f@kernel.org>
From:   Alex Elder <elder@linaro.org>
In-Reply-To: <20221101213404.3e4c3b8f@kernel.org>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 11/1/22 11:34 PM, Jakub Kicinski wrote:
> On Sat, 29 Oct 2022 19:18:28 -0500 Alex Elder wrote:
>>   	/* Set up the defined endpoint bitmap */
>>   	ipa->defined = bitmap_zalloc(ipa->endpoint_count, GFP_KERNEL);
>>   	ipa->set_up = bitmap_zalloc(ipa->endpoint_count, GFP_KERNEL);
>> +	ipa->enabled = bitmap_zalloc(ipa->endpoint_count, GFP_KERNEL);
>>   	if (!ipa->defined || !ipa->set_up) {
> 
> This condition should now check if ipa->enabled

You're right.

> And the error handling patch needs to free it, in case it was something
> else that didn't get allocated?

See below, but in any case, I'll make sure this is right.

> Frankly I have gotten mass-NULL-checks wrong more than once myself so
> I'd steer clear of those, they are strangely error prone.

I don't typically do it, and generally don't like it,
but I think I was trying to make it look cleaner
somehow.  I'll check every one in separately in v2.

>>   		dev_err(dev, "unable to allocate endpoint bitmaps\n");
> 
> this error message should not be here (patch 5 adds it I think)
> memory allocation failures produce a splat, no need to print errors

At the end of the series, the structure of this
error handling changes, and I think this is
correct.  But now that you point this out I
think the way this evolves in the series could
use some improvement so I'll take another look
at it and hopefully make it better.

I don't normally report anything on allocation
failures for that reason; thanks for pointing
this out.

I really appreciate your feedback.  I'll send
out version 2 today.

					-Alex

>> +		bitmap_free(ipa->set_up);
>> +		ipa->set_up = NULL;
>>   		bitmap_free(ipa->defined);