-
Notifications
You must be signed in to change notification settings - Fork 1.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Accommodate 0735ecb33 in trace_dmu.h #7096
Conversation
I only have a very basic understanding of the code and have no idea whether this is the proper fix. |
Codecov Report
@@ Coverage Diff @@
## master #7096 +/- ##
==========================================
+ Coverage 75.35% 75.45% +0.09%
==========================================
Files 296 296
Lines 95715 95713 -2
==========================================
+ Hits 72128 72216 +88
+ Misses 23587 23497 -90
Continue to review full report at Codecov.
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you for the PR! Just one comment below.
include/sys/trace_dmu.h
Outdated
@@ -62,7 +62,7 @@ DECLARE_EVENT_CLASS(zfs_delay_mintime_class, | |||
__entry->tx_lastsnap_txg = tx->tx_lastsnap_txg; | |||
__entry->tx_lasttried_txg = tx->tx_lasttried_txg; | |||
__entry->tx_anyobj = tx->tx_anyobj; | |||
__entry->tx_waited = tx->tx_waited; | |||
__entry->tx_waited = tx->tx_dirty_delayed; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The printk
below looks like it's attempting to print out the fields of the dmu_tx_t
structure. I think you should rename tx_waited
to be tx_dirty_delayed
in the __entry
structure to be consistent.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, please update all the instances of tx_waited
in this macro to tx_dirty_delayed
.
include/sys/trace_dmu.h
Outdated
@@ -62,7 +62,7 @@ DECLARE_EVENT_CLASS(zfs_delay_mintime_class, | |||
__entry->tx_lastsnap_txg = tx->tx_lastsnap_txg; | |||
__entry->tx_lasttried_txg = tx->tx_lasttried_txg; | |||
__entry->tx_anyobj = tx->tx_anyobj; | |||
__entry->tx_waited = tx->tx_waited; | |||
__entry->tx_waited = tx->tx_dirty_delayed; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, please update all the instances of tx_waited
in this macro to tx_dirty_delayed
.
include/sys/trace_dmu.h
Outdated
@@ -73,7 +73,7 @@ DECLARE_EVENT_CLASS(zfs_delay_mintime_class, | |||
"anyobj %d waited %d start %llu wait_dirty %d err %i " |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
waited
should be replaced with dirty_delayed
in the format string as well.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
one more nit, and IMHO both commits should be squashed into one ;)
This change was missed in 0735ecb. Signed-off-by: András Korn <[email protected]>
@behlendorf (or whoever is responsible for the buildbot worker configs ;)) - it might make sense to configure one of the builders to have the appropriate kernel config/headers to compile with tracing support (see config/kernel-declare-event-class.m4), to prevent future changes that break it from slipping through like this one did? |
@Fabian-Gruenbichler actually we did. Unfortunately, we accidentally broke that builder in such a way that it was reporting success even though this was broken. @dinatale2 has opened openzfs/zfs-buildbot#128 to resolve that infrastructure issue. We'll get the bots updates once we sort out this issue and verify there aren't other issues which snuck in. Sorry about that. |
This change was missed in 0735ecb. Reviewed-by: Fabian Grünbichler <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: András Korn <[email protected]> Closes openzfs#7096
This change was missed in 0735ecb. Reviewed-by: Fabian Grünbichler <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: András Korn <[email protected]> Closes openzfs#7096
This change was missed in 0735ecb. Reviewed-by: Fabian Grünbichler <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: András Korn <[email protected]> Closes #7096
0735ecb renamed tx_waited to tx_dirty_delayed, but trace_dmu.h wasn't updated.
This is a minimal fix needed to enable the code to build, avoiding the error
include/sys/trace_dmu.h:65:31: error: ‘dmu_tx_t {aka struct dmu_tx}’ has
no member named ‘tx_waited’
__entry->tx_waited = tx->tx_waited;
^
Signed-off-by: András Korn [email protected]
Description
Motivation and Context
How Has This Been Tested?
It builds successfully on latest Debian sid, with the stock Debian kernel. It didn't build before.
I don't know how I could test the functionality of the impacted code.
Types of changes
Checklist:
Signed-off-by
.