July 2022 - Devel - Nfs Ganesha List Archives

[S] Change in ...nfs-ganesha[next]: STATE: When we have a grant cookie we need an obj reference

by Frank Filz (GerritHub)

Frank Filz has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541254 ) Change subject: STATE: When we have a grant cookie we need an obj reference ...................................................................... STATE: When we have a grant cookie we need an obj reference The way cookie management works, we could wind up needing to release the state lock but no longer have any obj references which would cause a use after free. Add apprpriate get_ref and put_ref to guard against this. Change-Id: Ifad01b1ab8bce23195d23aff77043b5b835026ca Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com> --- M src/SAL/state_lock.c 1 file changed, 28 insertions(+), 4 deletions(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/54/541254/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541254 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: Ifad01b1ab8bce23195d23aff77043b5b835026ca Gerrit-Change-Number: 541254 Gerrit-PatchSet: 1 Gerrit-Owner: Frank Filz <ffilzlnx(a)mindspring.com> Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

[M] Change in ...nfs-ganesha[next]: config: Skip mdcache_param init in cacheinode block read.

by Name of user not set (GerritHub)

prabhu.murugeshan(a)ibm.com has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541178 ) Change subject: config: Skip mdcache_param init in cacheinode block read. ...................................................................... config: Skip mdcache_param init in cacheinode block read. cacheinode and mdcache block uses common structure to read configfile. We don't need to initialize mdcache_param in both cacheinode and mdcache, just one time initialization is enough. Second time initialization wipe of setting read by other block. Changed mdcache and cacheinode order to give more preference to mdcache. Change-Id: If08c7b8f60ac4ded662ad96b18fb85be931a2cfe Signed-off-by: Prabhu Murugesan <prabhu.murugeshan(a)ibm.com> --- M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_read_conf.c M src/config_parsing/config_parsing.c 2 files changed, 71 insertions(+), 31 deletions(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/78/541178/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541178 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: If08c7b8f60ac4ded662ad96b18fb85be931a2cfe Gerrit-Change-Number: 541178 Gerrit-PatchSet: 1 Gerrit-Owner: prabhu.murugeshan(a)ibm.com Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

[S] Change in ...nfs-ganesha[next]: STATE: Fix ganesha crash at nlm4_unlock() in blocked lock req

by Name of user not set (GerritHub)

prabhu.murugeshan(a)ibm.com has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541176 ) Change subject: STATE: Fix ganesha crash at nlm4_unlock() in blocked lock req ...................................................................... STATE: Fix ganesha crash at nlm4_unlock() in blocked lock req When a blocked lock is granted, we need to take ref of file handle obj if the lock list is empty. The ref will be released in unlock() if the lock list is empty. This will fix possible crash in blocked lock request. With out this fix, the ganesha will crash at nlm4_unlock() if a lock is granted to client after lock conflict and the file is removed by another thread from same client. Change-Id: Idaee9e85574969de089950fac3e50ae736a170de Signed-off-by: Prabhu Murugesan <prabhu.murugeshan(a)ibm.com> --- M src/SAL/state_lock.c 1 file changed, 25 insertions(+), 0 deletions(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/76/541176/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541176 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: Idaee9e85574969de089950fac3e50ae736a170de Gerrit-Change-Number: 541176 Gerrit-PatchSet: 1 Gerrit-Owner: prabhu.murugeshan(a)ibm.com Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

[S] Change in ...nfs-ganesha[next]: STATE: Fix blocked lock lock_entry refcount

by Frank Filz (GerritHub)

Frank Filz has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541117 ) Change subject: STATE: Fix blocked lock lock_entry refcount ...................................................................... STATE: Fix blocked lock lock_entry refcount When a blocked lock is granted, we process that on a fridgethread. To do this, we need to hold a lock_entry reference (which protects the fsal_obj_handle and the export also) from after we schedule (before we release the state_lock) to after we are done processing. Change-Id: Ic67b0fed4b654d051d9664fe337f2207f3ffa066 Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com> --- M src/SAL/state_lock.c 1 file changed, 33 insertions(+), 1 deletion(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/17/541117/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541117 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: Ic67b0fed4b654d051d9664fe337f2207f3ffa066 Gerrit-Change-Number: 541117 Gerrit-PatchSet: 1 Gerrit-Owner: Frank Filz <ffilzlnx(a)mindspring.com> Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

[S] Change in ...nfs-ganesha[next]: CMakefiles: specific the Python to Python3

by vicente (GerritHub)

vicente has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541028 ) Change subject: CMakefiles: specific the Python to Python3 ...................................................................... CMakefiles: specific the Python to Python3 - That make system to know we need python3 instead python2. Because the python scripts are already base on python3 we could check python3 for CMakeFiles. - Also remove the Development checking because we do not use it. Change-Id: I0cd0ff0350a8db2c4822835930873ccbd6245444 Signed-off-by: Vicente Cheng <vicente.cheng(a)suse.com> --- M src/CMakeLists.txt M src/scripts/ganeshactl/CMakeLists.txt M src/scripts/gpfs-epoch/CMakeLists.txt 3 files changed, 28 insertions(+), 12 deletions(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/28/541028/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541028 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: I0cd0ff0350a8db2c4822835930873ccbd6245444 Gerrit-Change-Number: 541028 Gerrit-PatchSet: 1 Gerrit-Owner: vicente <freeze.bilsted(a)gmail.com> Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

[M] Change in ...nfs-ganesha[next]: spec: correct the related path on nfs-ganesha.spec-in.cmake

by vicente (GerritHub)

vicente has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541027 ) Change subject: spec: correct the related path on nfs-ganesha.spec-in.cmake ...................................................................... spec: correct the related path on nfs-ganesha.spec-in.cmake - Also the python3.6 is the also avaiable on centos7. Maybe we could use python3 on centos7. Change-Id: Ib0f2bfd82aa1f14fb6cede817c39d77632086129 Signed-off-by: Vicente Cheng <vicente.cheng(a)suse.com> --- M src/nfs-ganesha.spec-in.cmake 1 file changed, 46 insertions(+), 32 deletions(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/27/541027/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541027 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: Ib0f2bfd82aa1f14fb6cede817c39d77632086129 Gerrit-Change-Number: 541027 Gerrit-PatchSet: 1 Gerrit-Owner: vicente <freeze.bilsted(a)gmail.com> Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

[L] Change in ...nfs-ganesha[next]: ganesha-top: add tool named ganesha-top

by vicente (GerritHub)

vicente has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541001 ) Change subject: ganesha-top: add tool named ganesha-top ...................................................................... ganesha-top: add tool named ganesha-top ganesha-top is the tool that could monitoring that Process/OP/Client/Export general information like top. - Also add README.md Change-Id: I10bfdeb9015973de6b9d11b81cf84c669b77477a Signed-off-by: Vicente Cheng <vicente.cheng(a)suse.com> --- A src/ganesha_top/README.md A src/ganesha_top/ganesha-top 2 files changed, 648 insertions(+), 0 deletions(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/01/541001/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/541001 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: I10bfdeb9015973de6b9d11b81cf84c669b77477a Gerrit-Change-Number: 541001 Gerrit-PatchSet: 1 Gerrit-Owner: vicente <freeze.bilsted(a)gmail.com> Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

[M] Change in ...nfs-ganesha[next]: Trying to reap the MDCache LRU entries traversing through the whole q...

by Sagar Singh (GerritHub)

Sagar Singh has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/540965 ) Change subject: Trying to reap the MDCache LRU entries traversing through the whole queue, instead of checking only the first LRU entry of every lane. ...................................................................... Trying to reap the MDCache LRU entries traversing through the whole queue, instead of checking only the first LRU entry of every lane. When lru entries used are greater than high watermark then lru_reap_impl() try to reap the first entry of LRU queue for given qid, going through every lane. If the refcount is not (LRU_SENTINEL_REFCOUNT + 1), then we are leaving the whole lane, in which There are many entries which can be reaped. So this is slowing down the reaping and causing high memory use by the ganesha server. This scenario was resulting to run OOM Killer, so ganesha server was crashing with a segmentation fault. This patch is written in the hope that fast LRU reaping will not cause high memory usage by the MDCache. Change-Id: I2a2fae830ce678b8004ac80aa648b6cbffab77cf Signed-off-by: Sagar Singh <sagar.it.iiests(a)gmail.com> --- M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.h 2 files changed, 135 insertions(+), 1 deletion(-) git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/65/540965/1 -- To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/540965 To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings Gerrit-Project: ffilz/nfs-ganesha Gerrit-Branch: next Gerrit-Change-Id: I2a2fae830ce678b8004ac80aa648b6cbffab77cf Gerrit-Change-Number: 540965 Gerrit-PatchSet: 1 Gerrit-Owner: Sagar Singh <sagar.it.iiests(a)gmail.com> Gerrit-MessageType: newchange

2 years, 5 months

1
0
0 / 0

2024

2023

2022

2021

2020

2019

2018

Devel July 2022