netdev - Re: [PATCH net-next 02/14] net: cdc_ncm: use device rx

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:	Fri, 19 Oct 2012 12:01:39 +0200
From:	Alexey Orishko <alexey.orishko@...il.com>
To:	Bjørn Mork <bjorn@...k.no>
Cc:	Oliver Neukum <oliver@...kum.org>, netdev@...r.kernel.org,
	linux-usb@...r.kernel.org,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	Greg Suarez <gpsuarez2512@...il.com>,
	"Fangxiaozhi (Franko)" <fangxiaozhi@...wei.com>,
	Dan Williams <dcbw@...hat.com>,
	Aleksander Morgado <aleksander@...edo.com>
Subject: Re: [PATCH net-next 02/14] net: cdc_ncm: use device rx_max value if
 update failed

On Fri, Oct 19, 2012 at 11:30 AM, Bjørn Mork <bjorn@...k.no> wrote:
> Bjørn Mork <bjorn@...k.no> writes:
>> Alexey Orishko <alexey.orishko@...il.com> writes:
>>
> OK, I did some more experiments, and I am wondering if the real firmware
> problem is in the MBIM descriptor.  It is
>
>       CDC MBIM:
>         bcdMBIMVersion       1.00
>         wMaxControlMessage   1536
>         bNumberFilters       16
>         bMaxFilterSize       40
>         wMaxSegmentSize      4096
>         bmNetworkCapabilities 0x20
>           8-byte ntb input size
>
>
> so we use the 8-byte version of USB_CDC_SET_NTB_INPUT_SIZE, which fails
> with -EPIPE.  But forcing the 4-byte version seems to work. Hmm, I also
> see that the device returns 4 bytes in response to at
> USB_CDC_GET_NTB_INPUT_SIZE with an 8-byte buffer.  Maybe we can
> auto-quirk based on that?  I.e., if USB_CDC_GET_NTB_INPUT_SIZE returns
> only 4 bytes then we assume that the bmNetworkCapabilities flag is
> wrong.
>
> Is that acceptable?  Then it seems we are able to inform this device of
> the reduced buffer, and the other problems can be ignored.
>

Based on NCM errata, NCM functional descriptor: if Bit 5 is set, then
device can (must) handle 8-byte form of Set/GetNtbInputSize.
If they set a flag, but don't support the feature, hm.. is it a
prototype device or
commercially available one?

At least device must support Set request, but for GetNtbInputSize we
could happily
live without wNtbInMaxDatagrams (i.e. use 4 byte variant) on Linux.
Since we must anyway receive a complete NTB and making any number of skb
buffers from received NTB is not a problem at all.

regards,
alexey
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html