linux/drivers/block, branch v2.6.12

When cfq I/O scheduler is selected, get_request() in __make_request() calls

2005-06-17T14:15:10Z

__cfq_get_queue(). __cfq_get_queue() finds an existing queue (struct cfq_queue) of the current process for the device and returns it. If it's not found, __cfq_get_queue() creates and returns a new one if __cfq_get_queue() is called with __GFP_WAIT flag, or __cfq_get_queue() returns NULL (this means that get_request() fails) if no __GFP_WAIT flag. On the other hand, in __make_request(), get_request() is called without __GFP_WAIT flag at the first time. Thus, the get_request() fails when there is no existing queue, typically when it's called for the first I/O request of the process to the device. Though it will be followed by get_request_wait() for general case, __make_request() will just end the I/O with an error (EWOULDBLOCK) when the request was for read-ahead. Signed-off-by: Jens Axboe Signed-off-by: Kiyoshi Ueda Signed-off-by: Jun'ichi Nomura

This patch fixes q->unplug_thresh condition check in

2005-06-16T10:57:31Z

__elv_add_request(). rq.count[READ] + rq.count[WRITE] can increase more than one if another thread has allocated a request after the current request is allocated or in_flight could have changed resulting in larger-than-one change of nrq, thus breaking the threshold mechanism. Signed-off-by: Jens Axboe Signed-off-by: Tejun Heo

This patch kills elevator_global_init() in elevator.c which does

2005-06-16T10:56:15Z

nothing. Signed-off-by: Jens Axboe Signed-off-by: Tejun Heo

[PATCH] USB: fix ub issues

2005-06-09T08:38:11Z

This smoothes two imperfections: - Increase number of LUNs per device from 4 to 9. The best solution would be to remove this limit altogether, but that has to wait until the time when more than 26 hosts are allowed. - Replace mdelay with msleep in a probing routine. Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman

[PATCH] USB: Support multiply-LUN devices in ub

2005-06-03T07:04:27Z

Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman diff -urp -X dontdiff linux-2.6.12-rc3/drivers/block/ub.c linux-2.6.12-rc3-lem/drivers/block/ub.c

[PATCH] packet driver permission checking fix

2005-05-20T22:48:16Z

If you tried to open a packet device first in read-only mode and then a second time in read-write mode, the second open succeeded even though the device was not correctly set up for writing. If you then tried to write data to the device, the writes would fail with I/O errors. This patch prevents that problem by making the second open fail with -EBUSY. Signed-off-by: Peter Osterlund Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

[PATCH] CDRW/DVD packet writing data corruption fix

2005-05-17T14:59:18Z

I found a bug in the packet writing driver that could cause data corruption. The problem arised if the driver got a write request for a sector in a "zone" it was already working on. In that case it was supposed to queue the write request until it was done processing earlier requests for the same zone, and instead work on some other zone in the mean time. However, if there was no other zone to work on, the driver would initiate two packet_data objects for the same zone, causing unpredictable things to happen. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

[PATCH] Fix root hole in pktcdvd

2005-05-17T04:07:31Z

ioctl_by_bdev may only be used INSIDE the kernel. If the "arg" argument refers to memory that is accessed by put_user/get_user in the ioctl function, the memory needs to be in the kernel address space (that's the set_fs(KERNEL_DS) doing in the ioctl_by_bdev). This works on i386 because even with set_fs(KERNEL_DS) the user space memory is still accessible with put_user/get_user. That is not true for s390. In short the ioctl implementation of the pktcdvd device driver is horribly broken. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

[PATCH] Fix root hole in raw device

2005-05-17T04:07:21Z

[Patch] Fix raw device ioctl pass-through Raw character devices are supposed to pass ioctls through to the block devices they are bound to. Unfortunately, they are using the wrong function for this: ioctl_by_bdev(), instead of blkdev_ioctl(). ioctl_by_bdev() performs a set_fs(KERNEL_DS) before calling the ioctl, redirecting the user-space buffer access to the kernel address space. This is, needless to say, a bad thing. This was noticed first on s390, where raw IO was non-functioning. The s390 driver config does not actually allow raw IO to be enabled, which was the first part of the problem. Secondly, the s390 kernel address space is distinct from user, causing legal raw ioctls to fail. I've reproduced this on a kernel built with 4G:4G split on x86, which fails in the same way (-EFAULT if the address does not exist kernel-side; returns success without actually populating the user buffer if it does.) The patch below fixes both the config and address-space problems. It's based closely on a patch by Jan Glauber , which has been tested on s390 at IBM. I've tested it on x86 4G:4G (split address space) and x86_64 (common address space). Kernel-address-space access has been assigned CAN-2005-1264. Signed-off-by: Stephen Tweedie Signed-off-by: Dave Jones Signed-off-by: Greg Kroah-Hartman

[PATCH] drivers/block/rd.c: rd_size shouldn't be static

2005-05-06T23:58:18Z

I somehow missed that there is external usage of rd_size on some architectures. Signed-off-by: Adrian Bunk Signed-off-by: Linus Torvalds