linux-kernel - Re: [PATCH v3]ipmi:bt-bmc:Avoid unnecessary judgement

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:   Mon, 13 Apr 2020 16:59:41 -0500
From:   Corey Minyard <minyard@....org>
To:     Tang Bin <tangbin@...s.chinamobile.com>
Cc:     arnd@...db.de, gregkh@...uxfoundation.org,
        openipmi-developer@...ts.sourceforge.net,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3]ipmi:bt-bmc:Avoid unnecessary judgement

On Mon, Apr 13, 2020 at 11:44:49PM +0800, Tang Bin wrote:
> Hi Corey:
> 
> On 2020/4/13 22:23, Corey Minyard wrote:
> > > Can I consider that the patch will be applied in 5.8？
> > It's in my queue, so that's the plan.
> > 
> > > >    I
> > > > changed the title to be "Avoid unnecessary check".
> > > You have modified it, which means I don't need to submit a new patch？
> > Correct.
> 
> Thank you very much, I am waiting for the applied.
> 
> 
> Then, I have some questions to ask you:
> 
>     I have checked the file bt-bmc.c carefully, and found that there are
> another two problems.Please help me analyze them, if you think it is
> feasible, then I will submit the patch.
> 
>     Q1: About Format Problem
> 
>            In the 469~471 line, the first letter should be indented, please
> check if the writing here is reasonable?
> 

I'm not sure how that happened.  It was that way from the original
submitter and nobody noticed in review.  It's obviously wrong.

> 
>     Q2: About the function bt_bmc_config_irq()
> 
>           1）In the function bt_bmc_probe(), the return value of
> bt_bmc_config_irq() made no judgement, whether it is suitable? （If your
> view is don't need to judge, the following will change.）
> 

Hmm, that's probably not a big deal.  If it fails irqs are just not
used.  It probably shouldn't return a value, though.

> 
>           2）According to the kernel interface of platform_get_irq(),the
> return value is negative,
> 
>                    if (!bt_bmc->irq)
>                         return -ENODEV;
> 
>                so the check here is invalid.The standard way to write is:
> 
>                      if (bt_bmc->irq < 0)
>                           return bt_bmc->irq;
> 
>                But consider if failed, "bt_bmc->irq" must be assigned to
> "0"，the easiest way is to delete the        403~404 line, handled directly
> by the function devm_request_irq().

The problem you point out is real, the check should be < 0.

You don't want it handled by devm_request_irq, that could result in logs
that are invalid.

Also, it should use platform_get_irq_optional() to avoid a spurrious log
when there is no irq.

> 
> 
>         Q3：About dev_warm()
> 
>                 KERN_WARNING is higher than KERN_INFO, the same to
> dev_warn() and dev_info(). When the function bt_bmc_probe() uses dev_info()
> to print error message, the dev_warm() in the line of 409 should be
> redundant.

That is all correct as it is.  If there is an irq specified and it can't
be requested, that is a problem.  If there is no irq specified, that is
fine, just info is good.

Thanks,

-corey

> 
> 
> I am waiting for your replay, and thank you for your guidance.
> 
> Tang Bin
> 
> 
>