- 17 Aug, 2014 17 commits
-
-
Konrad Zapalowicz authored
The polling counter variable is only exposed via sysfs and has no other purpose. Now, since the polling shall be implemented as a board specific feature rather than being global in the driver this counter is obsolete. Signed-off-by:
Konrad Zapalowicz <bergo.torino@gmail.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Konrad Zapalowicz authored
This commit removes the driver's global state variable. This is ok because the state was changed only once at the end of init phase thus the future usage of this variable is pointless. Signed-off-by:
-
Purnendu Kapadia authored
we should use __packed attribute Signed-off-by:
Purnendu Kapadia <pro8linux@gmail.com> Signed-off-by:
-
Jeshwanth Kumar N K authored
Fixed up warnings, break is not useful after return statement. And the exit Label is deleted, now returning inline. Signed-off-by:
Jeshwanth Kumar N K <jeshkumar555@gmail.com> Signed-off-by:
-
Hema Prathaban authored
This patch fixes the following warning using checkpatch.pl WARNING: Prefer [subsystem eg: netdev]_warn([subsystem]dev, ... then dev_warn(dev, ... then pr_warn(... to printk(KERN_WARNING ... Signed-off-by:
Hema Prathaban <hemaklnce@gmail.com> Signed-off-by:
-
Hema Prathaban authored
This patch fixes the following error using checkpatch.pl Error: do not use assignment in if condition Signed-off-by:
-
Oleg Drokin authored
We have a bunch of typedefs for common things that made no sense and hid the actual type from plain view. Replace them with proper uXX or sXX types. Exception is in lustre_idl.h where they are replaced with __uXX and __sXX to be able to be included in userspace Signed-off-by:
Oleg Drokin <green@linuxhacker.ru> Signed-off-by:
-
Oleg Drokin authored
seqno_t is u64 and mdsno_t is u32 so just use them as such. Signed-off-by:
-
Joe Perches authored
DBG_PRT uses are unnecessarily complex. Convert DBG_PRT msglevel to pr_<level>. This changes the KERN_<level> type of several uses. It also enables dynamic_debug for the pr_debug conversions. This patch can be a prelude to converting these pr_<level> uses to dev_<level> as appropriate. Other changes: Realign arguments of these conversions. Remove now unused static int msglevel declarations. Remove now unused DBG_PRT #define. Compile tested only. Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Oleg Drokin authored
Reported by Dan Carpenter <dan.carpenter@oracle.com> outsize = sizeof(*gfout) + gfin->gf_pathlen; Where outsize is int and gf_pathlen is u32 from userspace can lead to integer overflowwhere outsize is some small number less than sizeof(*gfout) Add a check for pathlen to be of sensical size. Signed-off-by:
Oleg Drokin <oleg.drokin@intel.com> Reviewed-on: http://review.whamcloud.com/11412 Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5476Reviewed-by:
Dmitry Eremin <dmitry.eremin@intel.com> Reviewed-by:
John L. Hammond <john.hammond@intel.com> Signed-off-by:
-
Frank Zago authored
The only parameter from userspace that matters is the length of the buffer. We don't need to allocate then import the whole structure. By importing only that length, we can save a memory allocation. Add sparse annotations to that function. Signed-off-by:
frank zago <fzago@cray.com> Reviewed-on: http://review.whamcloud.com/11167 Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5389Reviewed-by:
Andreas Dilger <andreas.dilger@intel.com> Signed-off-by:
-
John L. Hammond authored
ll_setattr_raw() expects to be called with the inode mutex held so do when calling it from ll_hsm_import(). Signed-off-by:
Jinshan Xiong <jinshan.xiong@intel.com> Reviewed-by:
Faccini Bruno <bruno.faccini@intel.com> Signed-off-by:
-
John L. Hammond authored
In cl_req_alloc() ensure that the list heads crq_pages and crq_layers have been initialized before passing the request to cl_req_completion(). This fixes an oops in the error path. Signed-off-by:
Bobi Jam <bobijam@gmail.com> Reviewed-by:
-
Fan Yong authored
In very old release (older than Lustre-1.8), if the client talks with the server that does not support ibits lock, then the client needs to convert it as plain lock. Such interoperability check and convertion is out of date for a long time. Drop it. Signed-off-by:
Fan Yong <fan.yong@intel.com> Reviewed-on: http://review.whamcloud.com/11004 Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4971Reviewed-by:
wangdi <di.wang@intel.com> Signed-off-by:
-
Paul Cassella authored
In ll_fault0, the 'fault' struct is mostly cleared before the call to cl_io_loop, but ft_flags is not reset. It is ordinarily set by the call to filemap_fault in vvp_io_kernel_fault, but if Lustre returns before calling filemap_fault, it still has the old value of ft_flags. ll_fault0 will then consume the ft_flags field. If it has the VM_FAULT_RETRY bit set, it will be used as ll_fault0() and ll_fault()'s return value. This is a problem when VM_FAULT_RETRY is in ft_flags: When fault/filemap_fault return with that flag set, they have already released the mmap semaphore, and do_page_fault does not need to release it. Incorrectly returning this flag from ll_fault means mmap_sem is not upped in the kernel's do_page_fault(). In addition to clearing ft_flags, this patch does not use it unless it is valid. It's potentially misleading to return ft_flags in "fault_ret" if ft_flags has not been set by filemap_fault. This adds clarity, but does not change the current behavior: When not valid, ft_flags is replaced by fault_ret, which is zero, as is ft_flags when not set by filemap_fault. Signed-off-by:
Patrick Farrell <paf@cray.com> Reviewed-on: http://review.whamcloud.com/10956 Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5291Reviewed-by:
-
John L. Hammond authored
In mdc_enqueue() clear the lock handle, lock mode, and request stored in the lookup intent if mdc_finish_enqueue() fails. Signed-off-by:
Lai Siyao <lai.siyao@intel.com> Signed-off-by:
-
Nathaniel Clark authored
Check to make sure total size of request does not overflow when calculated. Return -1 from hur_len() if it does overflow. Signed-off-by:
Nathaniel Clark <nathaniel.l.clark@intel.com> Reviewed-on: http://review.whamcloud.com/10615 Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4984Reviewed-by:
-
- 16 Aug, 2014 23 commits
-
-
Artemiy Volkov authored
This patch fixes the 'quoted string split across lines' checkpatch.pl warning in prism2fw.c. Signed-off-by:
Artemiy Volkov <artemiyv@acm.org> Signed-off-by:
-
vibi sreenivasan authored
This patch fixes the coding style error : "WARNING: else is not generally useful after a break or return" reported by checkpatch.pl Signed-off-by:
Vibi Sreenivasan <vibisreenivasan@linuxmail.org> Signed-off-by:
-
H Hartley Sweeten authored
There are a couple comedi drivers that duplicate some of the register defines used by the 8255 module. Move these defines into the header so the duplication can be removed. Signed-off-by:
H Hartley Sweeten <hsweeten@visionengravers.com> Reviewed-by:
Ian Abbott <abbotti@mev.co.uk> Signed-off-by:
-
H Hartley Sweeten authored
The drivers that use this module with memory mapped io all have the ioremap'ed base address stored in the comedi_device 'mmio' member. Introduce a default (*io) function that does 8-bit memory mapped io. Modify subdev_8255_init() so that it takes a flag parameter indicating if the io is port or memory mapped. Make the function static and rename it to __subdev_8255_init(). Introduce two exported wrappers for __subdev_8255_init(): subdev_8255_init() - for drivers that do 8-bit port io subdev_8255_mm_init() - for drivers that do 8-bit memory mapped io Use subdev_8255_mm_init() in the drivers that do 8-bit memory mapped io and remove the private (*io) functions. Signed-off-by:
-
H Hartley Sweeten authored
The async command support in this module is incomplete and nothing has ever used it. Just remove it. Signed-off-by:
-
H Hartley Sweeten authored
Currently, all users of is module that use the default (*io) function pass an 'iobase' to subdev_8255_init() of the form: dev->iobase + OFFSET_TO_8255_BASE_REG Now that the (*io) callback includes the comedi_device 'dev' pointer the 'dev->iobase' does not need to be included. Modify the default (*io) function, subdev_8255_io(), to automatically add the dev->iobase to the address when reading/writing the port. For aesthetics, rename the subdevice private data member to 'regbase'. Also, rename the local variables in this module that are used to access this member. Add a comment in dev_8255_attach() about the 'iobase' that is passed to subdev_8255_init(). For manually attached 8255 devices the io region is requested with __comedi_request_region() which does not set dev->iobase. For these devices the 'regbase' is actually the 'iobase'. Remove the, now unnecessary, dev->iobase from all the callers of subdev_8255_init(). There are a couple drivers that only passed the dev->iobase. For those drivers pass a 'regbase' of 0x00. Note that the das16m1 driver is a bit goofy. The devpriv->extra_iobase is requested using __comedi_request_region() which does not set the dev->iobase. But the starting address passed is dev->iobase + DAS16M1_82C55 so a 'regbase' of DAS16M1_82C55 is passed to subdev_8255_init(). Signed-off-by:
-
H Hartley Sweeten authored
The 8255 driver (*io) callback now includes the comedi_device pointer. Using this we can get the ioremap'ed base address. Instead of passing the (cast) mmio address to subdev_8255_init(), pass the 'iobase' of the 8255 registers (i * 4). Signed-off-by:
-
H Hartley Sweeten authored
The 8255 driver (*io) callback now includes the comedi_device pointer. Using this we can get the ioremap'ed base address. Instead of passing the (cast) mmio address to subdev_8255_init(), pass the 'iobase' of the 8255 registers (I8255_4020_REG). Signed-off-by:
-
H Hartley Sweeten authored
The 8255 driver (*io) callback now includes the comedi_device pointer. Using this we can get the ioremap'ed base address. Instead of passing the (cast) mmio address to subdev_8255_init(), pass the 'iobase' of the 8255 registers (DIO_8255_OFFSET). Signed-off-by:
-
H Hartley Sweeten authored
The 8255 driver (*io) callback now includes the comedi_device pointer. Using this we can get the ioremap'ed base address. Instead of passing the (cast) mmio address to subdev_8255_init(), pass the 'iobase' of the 8255 registers (dioP2ExpansionIO8Bit). Signed-off-by:
-
H Hartley Sweeten authored
The 8255 driver (*io) callback now includes the comedi_device pointer. Instead of passing the (cast) pointer to subdev_8255_init(), pass the 'iobase' of the 8255 registers (DIO_BASE_REG). Signed-off-by:
-
H Hartley Sweeten authored
The 8255 driver (*io) callback now includes the comedi_device pointer. Instead of passing the (cast) pointer to subdev_8255_init(), pass the 'iobase' of the 8255 registers (Port_A). Signed-off-by:
-
H Hartley Sweeten authored
The 8255 driver uses an (*io) callback to read/write the registers of the 8255 device. The default callback provided by the driver uses inb()/outb() calls to access to registers based on an 'iobase' that was initialized during the subdev_8255_init() and a 'port' value. The users of this module can optionally provide a custom (*io) callback to handle the read/write in another manner. Make the (*io) callback a bit more flexible by also passing the comedi_device pointer as a parameter. Signed-off-by:
-
H Hartley Sweeten authored
This driver uses two iobase addresses, found in PCI bars 2 and 3. Currently, the address in PCI bar 2 is saved in the private data as 'iobase1' and the address in PCI bar 3 is saved in the comedi_device as the 'iobase'. The 'iobase' is the base address of the daq registers (ai/ao) of the board. The 'iobase1' address is the base address of the 8255, 8254, configuration, and interrupt registers. Flip the saving of these base addresses. Save the address from PCI bar 2 in the comedi_device 'iobase' and the address from PCI bar 3 in the private data as 'daqio'. This will help with some cleanup of the 8255 module. Signed-off-by:
-
H Hartley Sweeten authored
This driver uses two iobase addresses, found in PCI bars 2 and 3. Currently, the address in PCI bar 3 is saved in the comedi_device as the 'iobase' and the one in PCI bar 2 is just passed to subdev_8255_init() as the 'iobase' parameter. Save the PCI bar 3 address in the private data as 'daqio' and the address from PCI bar 2 in the comedi_device as the 'iobase'. This will help with some cleanup of the 8255 module. Signed-off-by:
-
H Hartley Sweeten authored
This driver uses three iobase addresses, found in PCI bars 2, 3, and 4. Currently, the address in PCI bar 2 is saved in the comedi_device as the 'iobase', the PCI bar 3 address is saved in the private data as 'BADR3' and the one in PCI bar 4 is just passed to subdev_8255_init() as the 'iobase' parameter. Flip the saving of the PCI bar 2 and 4 base addresses. Save the address from PCI bar 2 in the private data as the 'daqio' and the address from PCI bar 4 in the comedi_device as the 'iobase'. This will help with some cleanup of the 8255 module. Signed-off-by:
-
H Hartley Sweeten authored
Currently the base address of the 8254 and 8255 devices, found in PCI bar 3, is saved in the private data as 'pacer_counter_dio'. The 'iobase' in the comedi_device is currently unused. Save the address from PCI bar 3 in the comedi_device and remove the unnecessary member from the private data. This will help with some cleanup of the 8255 module. Signed-off-by:
-
H Hartley Sweeten authored
This exported function just does a free_irq() to release the interrupt handler for the legacy and PCI dio200 drivers. The legacy driver also calls comedi_legacy_detach() which would also do the free_irq(). For that driver the just use comedi_legacy_detach() directly for the (*detach). For the PCI driver, add the free_irq() to the private (*datach) function. Remove the, then unused, dio200_common_detach() function. Signed-off-by:
-
H Hartley Sweeten authored
This function is only called by amplc_dio200_common_attach() and it can never fail. For aesthetics, absorb it into that function. Signed-off-by:
-
H Hartley Sweeten authored
This member of the boardinfo is only set for the PCIE boards. Use the 'is_pcie' flag to determine if the offset needs to be shifted when reading/writing the registers. Signed-off-by:
-
H Hartley Sweeten authored
This member of the boardinfor is only set for the PCIE boards. For aeshetics, rename it to 'is_pcie'. For clarity, use this flag in the (*auto_attach) to determine if the dio200_pcie_board_setup() function needs to be called instead of using the switch (context_model). Signed-off-by:
-
H Hartley Sweeten authored
The function names provide enough description. The extra comments are not necessary. Remove them. Also, tidy up some of the function declarations. Signed-off-by:
-
H Hartley Sweeten authored
This struct is used to provide part of the boardinfo data. Using the extra indirection does not provide any additional clarity to the driver. Absorb the members from dio200_layout into dio200_board and remove the extra 'layout' indirection. For aesthetics, rename all the local variables used for the boardinfo pointer to 'board'. Signed-off-by:
-
