Re: [PATCH 1/3] usb/storage: a couple defines from drivers/usb/storage/transport.h to include/linux/usb/storage.h

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



* Greg KH | 2012-02-25 09:33:44 [-0800]:

>> Now that I looked at the ub.c driver I feel like removing more stuff. On
>> the first look, it seems to do same thing as the storage driver in
>> drivers/usb/storage. It is not enable on Debian sid box and it is also
>> not enabled on my FC15 box. This makes me ask the following question:
>> May I remove it or are there some users which can not use the storage
>> driver?
>
>There might still be some users, so it should stay, it's not harming
>anyone as-is, right?

I googled a little and I haven't any bug reports saying "use ub, the
usb-storage driver can't handle it" but I found some reports which said
"disable ub and try again". Those reports were old, like from 2004 so I
wouldn't account them under harm.

Additionally to the distros I mentioned, it seems that ubuntu also
disabled this driver since Karmic so the "major" distros are not
shipping it. This leaves us with people that compile their own kernel.
Looking at the git history of ub.c I see only changes related to
clean ups of the block layer or some other parts of the kernel
which requires changing ub.c as well. I see a reset fix from "Mon Nov 10
21:11:11 2008 -0700" as the last "real" change. More than 3 years
without a change is okay for mature & working driver.
This driver does a lot on its own and does not use libusual. For
instance there is a workaround for READ CAPACITY which subtracts 1 LBA
from the number of reported block and I don't see this fix in ub.
Alan recently added a workaround ("usb-storage: redo incorrect reads")
which re-redoes the first READ10 command. I don't see such a thing added
to ub. The ub driver checks whether the signature in CSW is always the
same. This is also done by the usb-storage driver, however it has a
US_FL_BULK_IGNORE_TAG flag where it skips this check and some devices
seem to send bad & different signature. So those devices don't work with
ub at all.

This looks to me like there are some broken devices which work
usb-storage but don't with ub. Therefore I suggest to add this driver
to the removal list, add a prinkt that this driver will be removed. _IF_
there are people which rely on it and can not use usb-storage then they
should speak up and explain why.

After all I don't see a single reason why we should keep two different
drivers doing the same thing where one is smaller and more complete than
the other.

>greg k-h

Sebastian
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux