- 20 Feb, 2016 40 commits
-
-
Oleg Drokin authored
This fixes checkpatch messages about "spaces preferred around that '-' (ctx:VxV)" Signed-off-by:
Oleg Drokin <green@linuxhacker.ru> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Oleg Drokin authored
Found with checkpatch Signed-off-by:
-
Oleg Drokin authored
Adjust alignment of argments that were pushed to next lines to conform to kernel code style. Found with checkpatch. Signed-off-by:
-
Oleg Drokin authored
Checkpatch highlighted that some of our Lustre files carry this extra paragraph and indeed it does seem somewhat redundant, so remove it. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also remove some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" The only exception is ll_update_inode where dropping != NULL in the construction below would break the logic. I guess we could change lsm != NULL to !!lsm, but that's uglier. (lsm != NULL) == ((body->valid & OBD_MD_FLEASIZE) != 0) Also removed some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also removed some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
All instances of "x == NULL" are changed to "!x" and "x != NULL" to "x" Also removed some redundant assertions. Signed-off-by:
-
Oleg Drokin authored
The complicated version of osc_page_protected and osc_page_is_dlocked are unsafe and were commented out for ages, so probably no point in carrying them on. Signed-off-by:
-
Oleg Drokin authored
lock_lock_multi_match stayed commented out unused for ages now, so let's just remove it. Signed-off-by:
-
Oleg Drokin authored
class_get_type has some references to various server modules that could not exist on the server, so get rid of them. Signed-off-by:
-
Oleg Drokin authored
In client_obd_setup references to LUSTRE_OSP_NAME could only happen on metadata servers, so remove them as never true Signed-off-by:
-
Oleg Drokin authored
This only makes sense on metadata server, so get rid of it. Also remove now unused MDS_OSC_MAX_RIF_DEFAULT define Signed-off-by:
-
Oleg Drokin authored
All the users seems to have disappeared. Signed-off-by:
-
Oleg Drokin authored
Now that the sysfs conversion is complete, also convert all the remaining comments Signed-off-by:
-
Oleg Drokin authored
Now that the sysfs conversion is complete, also convert all the remaining comments Signed-off-by:
-
Oleg Drokin authored
Now that the sysfs conversion is complete, also convert all the remaining comments Signed-off-by:
-
Oleg Drokin authored
Now that the sysfs conversion is complete, also convert all the remaining comments Signed-off-by:
-
Oleg Drokin authored
Now that the sysfs conversion is complete, also convert all the remaining comments Signed-off-by:
-
Amir Shehata authored
This is the third patch of a set of patches that enables DLC. This patch adds the following features to LNET. Currently these features are not driven by user space. - Adding/Deleting Networks dynamically Two new functions were added: - lnet_dyn_add_ni() add an NI. if the NI is already added then fail with appropriate error code - lnet_dyn_del_ni() delete an existing NI. If NI doesn't exist fail with appropriate failure code. These functions shall be called from IOCTL. Signed-off-by:Amir Shehata <amir.shehata@intel.com> Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2456 Reviewed-on: http://review.whamcloud.com/9832Reviewed-by:
James Simmons <uja.ornl@gmail.com> Reviewed-by:
Doug Oucharek <doug.s.oucharek@intel.com> Reviewed-by:
Liang Zhen <liang.zhen@intel.com> Reviewed-by:
Oleg Drokin <oleg.drokin@intel.com> Signed-off-by:
-
Amir Shehata authored
This is the second patch of a set of patches that enables DLC. This patch adds the following features to LNET. Currently these features are not driven by user space. - Enabling Routing on Demand. The default number of router buffers are allocated. - Disable Routing on demand. Unused router buffers are freed and used router buffers are freed when they are no longer in use. The following time routing is enabled the default router buffer values are used. It has been decided that remembering the user set router buffer values should be remembered and re-set by user space scripts. - Increase the number of router buffers on demand, by allocating new ones. - Decrease the number of router buffers. Exccess buffers are freed if they are not in use. Otherwise they are freed once they are no longer in use. Signed-off-by:
-
Amir Shehata authored
This is the first patch of a set of patches that enables DLC. This patch adds some cleanup in the config.c as well as some preparatory changes in peer.c to enable dynamic network configuration Signed-off-by:
-
Frank Zago authored
LIBCFS_ALLOC already zero out the memory allocated, so there is no need to zero out the memory again. Signed-off-by:
Frank Zago <fzago@cray.com> Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5304 Reviewed-on: http://review.whamcloud.com/11012Reviewed-by:
Patrick Farrell <paf@cray.com> Reviewed-by:
Andreas Dilger <andreas.dilger@intel.com> Reviewed-by:
Dmitry Eremin <dmitry.eremin@intel.com> Reviewed-by:
-
Liang Zhen authored
returned value of lnet_post_send_locked and lnet_post_routed_recv_locked are changed to -ve by: http://review.whamcloud.com/#/c/9369/ this is wrong because callers rely on +ve to identify blocked message which is not a failure. To respect linux kernel coding style and not use positive error code, this patch adds two macros as non-error returned values of these functions: LNET_CREDIT_OK has credit for message LNET_CREDIT_WAIT no credit and message is blocked both these functions will return these two values instead of 0 and EAGAIN Signed-off-by:
Chris Horn <hornc@cray.com> Reviewed-by:
-
John L. Hammond authored
Remove most uses of IS_ERR_VALUE(). This macro was often given an int argument coming from PTR_ERR(). This invokes implementation defined behavior since the long value gotten by applying PTR_ERR() to a kernel pointer will usually not be representable as an int. Moreover it may be just plain wrong to do this since the expressions IS_ERR(p) and IS_ERR_VALUE((int) PTR_ERR(p)) are not equivalent for a general pointer p. Signed-off-by:
John L. Hammond <john.hammond@intel.com> Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3498 Reviewed-on: http://review.whamcloud.com/6759Reviewed-by:
-
James Simmons authored
Remove the bulk of code for the no longer supported SeaStar interconnect found on older Cray systems. Signed-off-by:
Chuck Fossen <chuckf@cray.com> Reviewed-by:
-
Liang Zhen authored
error handler of lnet_create_reply_msg() didn't release lnet_res_lock if lnet_msg_alloc() failed. It can be fixed by moving validation check of msg out from lock. Signed-off-by:
Isaac Huang <he.huang@intel.com> Reviewed-by:
-
Chris Horn authored
We consider routes "down" if the router is down or the router NI for the target network is down. This should be reflected in the output of /proc/sys/lnet/routes Signed-off-by:
Cory Spitz <spitzcor@cray.com> Reviewed-by:
-
Chris Horn authored
Rename several variables in lnet_find_route_locked to make the code easier to understand. Broken out of patch 7857. Signed-off-by:
-
