From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: Ingo Molnar <mingo@elte.hu>, LKML <linux-kernel@vger.kernel.org>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>,
Steven Rostedt <rostedt@goodmis.org>,
Thomas Gleixner <tglx@linutronix.de>,
Peter Zijlstra <a.p.zijlstra@chello.nl>
Subject: [RFC][PATCH 6/7] lockdep: Maintain rw_state entries in locklist
Date: Sun, 17 Apr 2011 11:45:11 +0200 [thread overview]
Message-ID: <20110417095507.318610186@chello.nl> (raw)
In-Reply-To: <20110417094505.865828233@chello.nl>
[-- Attachment #1: gautham_r_shenoy-lockdep-maintain_rw_state_entries_in.patch --]
[-- Type: text/plain, Size: 4216 bytes --]
From: Gautham R Shenoy <ego@in.ibm.com>
The dependencies are currently maintained using a structure named
locklist. For a dependency A --> B, it saves B's lock_class in an
entry that would be linked to A's locks_after list.
However, in order to make use of the split chains introduced in the
previous patch, we need to enhance this infrastructure to save the
read/write states of A and B for each dependency such that we might
distinguish between the read and write chains.
Signed-off-by: Gautham R Shenoy <ego@in.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
---
include/linux/lockdep.h | 6 ++++++
kernel/lockdep.c | 23 +++++++++++++++++------
2 files changed, 23 insertions(+), 6 deletions(-)
Index: linux-2.6/include/linux/lockdep.h
===================================================================
--- linux-2.6.orig/include/linux/lockdep.h
+++ linux-2.6/include/linux/lockdep.h
@@ -164,6 +164,10 @@ struct lockdep_map {
* lock_after/lock_before list of a particular lock.
* @dep_class - lock_class of the lock which is involved in a dependency with
* the lock to which this entry is linked to.
+ * @this_lock_rw_state - The read/write state of the lock to which this
+ * dependency entry belongs to.
+ * @dep_lock_rw_state - The read/write state of the lock with the lock class
+ * dep_class in this particular dependecy involvement.
*
* Every lock has a list of other locks that were taken after it.
* We only grow the list, never remove from it:
@@ -173,6 +177,8 @@ struct lock_list {
struct lock_class *dep_class;
struct stack_trace trace;
int distance;
+ unsigned int this_lock_rw_state:3;
+ unsigned int dep_lock_rw_state:3;
/*
* The parent field is used to implement breadth-first search, and the
Index: linux-2.6/kernel/lockdep.c
===================================================================
--- linux-2.6.orig/kernel/lockdep.c
+++ linux-2.6/kernel/lockdep.c
@@ -816,11 +816,13 @@ static struct lock_list *alloc_list_entr
/*
* Add a new dependency to the head of the list:
*/
-static int add_lock_to_list(struct lock_class *class, struct lock_class *this,
+static int add_lock_to_list(struct held_lock *this_hlock,
+ struct held_lock *dep_hlock,
struct list_head *head, unsigned long ip,
int distance, struct stack_trace *trace)
{
struct lock_list *entry;
+ struct lock_class *dep_class = hlock_class(dep_hlock);
/*
* Lock not present yet - get a new dependency struct and
* add it to the list:
@@ -829,9 +831,11 @@ static int add_lock_to_list(struct lock_
if (!entry)
return 0;
- entry->dep_class = this;
+ entry->dep_class = dep_class;
entry->distance = distance;
entry->trace = *trace;
+ entry->this_lock_rw_state = this_hlock->rw_state;
+ entry->dep_lock_rw_state = dep_hlock->rw_state;
/*
* Since we never remove from the dependency list, the list can
* be walked lockless by other CPUs, it's only allocation
@@ -1690,6 +1694,8 @@ check_prev_add(struct task_struct *curr,
if (entry->dep_class == hlock_class(next)) {
if (distance == 1)
entry->distance = 1;
+ entry->this_lock_rw_state |= prev->rw_state;
+ entry->dep_lock_rw_state |= next->rw_state;
return 2;
}
}
@@ -1697,19 +1703,24 @@ check_prev_add(struct task_struct *curr,
if (!trylock_loop && !save_trace(&trace))
return 0;
+ list_for_each_entry(entry, &hlock_class(next)->locks_before, entry) {
+ if (entry->dep_class == hlock_class(prev)) {
+ entry->this_lock_rw_state |= next->rw_state;
+ entry->dep_lock_rw_state |= prev->rw_state;
+ }
+ }
+
/*
* Ok, all validations passed, add the new lock
* to the previous lock's dependency list:
*/
- ret = add_lock_to_list(hlock_class(prev), hlock_class(next),
- &hlock_class(prev)->locks_after,
+ ret = add_lock_to_list(prev, next, &hlock_class(prev)->locks_after,
next->acquire_ip, distance, &trace);
if (!ret)
return 0;
- ret = add_lock_to_list(hlock_class(next), hlock_class(prev),
- &hlock_class(next)->locks_before,
+ ret = add_lock_to_list(next, prev, &hlock_class(next)->locks_before,
next->acquire_ip, distance, &trace);
if (!ret)
return 0;
next prev parent reply other threads:[~2011-04-17 9:58 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-04-17 9:45 [RFC][PATCH 0/7] lockdep: Support recurise-read locks Peter Zijlstra
2011-04-17 9:45 ` [RFC][PATCH 1/7] lockdep: Implement extra recursive-read lock tests Peter Zijlstra
2011-04-17 9:45 ` [RFC][PATCH 2/7] lockdep: Remove redundant read checks Peter Zijlstra
2011-04-18 14:28 ` Steven Rostedt
2011-04-17 9:45 ` [RFC][PATCH 3/7] lockdep: Annotate read/write states Peter Zijlstra
2011-04-18 13:34 ` Yong Zhang
2011-04-18 16:34 ` Steven Rostedt
2011-04-18 16:36 ` Peter Zijlstra
2011-04-18 16:26 ` Steven Rostedt
2011-04-18 16:27 ` Steven Rostedt
2011-04-18 16:31 ` Steven Rostedt
2011-04-17 9:45 ` [RFC][PATCH 4/7] lockdep: Seperate lock ids for read/write acquires Peter Zijlstra
2011-04-18 16:46 ` Steven Rostedt
2011-04-18 16:49 ` Peter Zijlstra
2011-04-18 17:33 ` Steven Rostedt
2011-04-18 22:07 ` Peter Zijlstra
2011-04-17 9:45 ` [RFC][PATCH 5/7] lockdep: Rename lock_list::class Peter Zijlstra
2011-04-17 9:45 ` Peter Zijlstra [this message]
2011-04-18 13:37 ` [RFC][PATCH 6/7] lockdep: Maintain rw_state entries in locklist Yong Zhang
2011-04-17 9:45 ` [RFC][PATCH 7/7] lockdep: Consider the rw_state of lock while validating the chain Peter Zijlstra
2011-04-18 3:41 ` [RFC][PATCH 0/7] lockdep: Support recurise-read locks Tetsuo Handa
2011-04-22 7:19 ` Yong Zhang
2011-04-22 7:27 ` Yong Zhang
2011-04-22 7:44 ` Tetsuo Handa
2011-04-22 8:01 ` Yong Zhang
2011-04-22 8:31 ` Tetsuo Handa
2011-04-22 8:59 ` Yong Zhang
2011-04-22 9:19 ` Tetsuo Handa
2011-04-23 12:33 ` [PATCH] lockdep: ignore cached chain key for recursive read Yong Zhang
2011-04-23 13:04 ` Tetsuo Handa
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20110417095507.318610186@chello.nl \
--to=a.p.zijlstra@chello.nl \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=penguin-kernel@I-love.SAKURA.ne.jp \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome