linux-kernel - Re: USB: free coherent buffer into atomic context

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-next>] [day] [month] [year] [list]

Message-ID: <AANLkTikj01UhkiVrGJg4jgoOtXZ6dJjnwTsskDUVLD_H@mail.gmail.com>
Date:	Sat, 18 Sep 2010 16:50:20 +0800
From:	Ming Lei <tom.leiming@...il.com>
To:	Wolfgang Mües <wolfgang@...w-muees.de>
Cc:	linux-usb <linux-usb@...r.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: USB: free coherent buffer into atomic context

CC linux-kernel

2010/9/18 Wolfgang Mües <wolfgang@...w-muees.de>:
> Hi,

Thanks for your reply.

>
> Am Freitag, 17. September 2010 16:59:25 schrieb Ming Lei:
>> The warning only means dma_free_coherent is called with irq disabled.
>> It is not very difficult to fix the warning in usb hcd, but I need to know
>>  why the WARN_ON below is added into dma_free_coherent.
>>
>>            WARN_ON(irqs_disabled());
>>
>> Anyone know the reason?
>
> Hmm... Coherent buffers are ment to be kept for the whole time a driver

Documentation/DMA_API.txt

        Consistent memory is memory for which a write by either the
        device or the processor can immediately be read by the processor
        or device without having to worry about caching effects.

IMO, it is not necessary for a coherent buffer to be kept for
the whole time a driver operates on a device. Also, coherent buffer
is very limited in some arch(such as ARM), maybe we should use it
more dynamically, instead of keeping it for long time.

> operates on a device. If such a buffer is freed in an interrupt context (== a

irqs_disabled() doesn't mean the buffer is freed in an interrupt context.

> momentary action), this is most likely a bug in the driver. At least it is a
> misuse of the concept of coherent buffers.

If it is only a misuse, seems WARN_ON can be removed.

Sorry, I still don't get the real reason why WARN_ON(irqs_disabled()) is
added in dma_free_coherent.

-- 
Lei Ming
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/