From: Peter Zijlstra <peterz@infradead.org>
To: Ingo Molnar <mingo@kernel.org>
Cc: rusty@rustcorp.com.au, mathieu.desnoyers@efficios.com,
oleg@redhat.com, paulmck@linux.vnet.ibm.com,
torvalds@linux-foundation.org, linux-kernel@vger.kernel.org,
andi@firstfloor.org, rostedt@goodmis.org, tglx@linutronix.de,
laijs@cn.fujitsu.com, linux@horizon.com
Subject: Re: [PATCH v5 10/10] module: Rework module_addr_{min,max}
Date: Tue, 14 Apr 2015 14:56:57 +0200 [thread overview]
Message-ID: <20150414125657.GM5029@twins.programming.kicks-ass.net> (raw)
In-Reply-To: <20150413165636.GH6040@gmail.com>
On Mon, Apr 13, 2015 at 06:56:36PM +0200, Ingo Molnar wrote:
> > + * Bounds of module allocation, for speeding up __module_address.
> > + * Protected by module_mutex.
> > + */
> > +static unsigned long module_addr_min = -1UL, module_addr_max = 0;
>
> I suspect the same .data vs. .bss problem affects the #else branch as
> well?
Yes, but the linear walk already has a 'problem', other than the linear
walk itself being one, the list_head isn't actually on the same line as
the 'key' entries -- although I suppose I could fix that for the
!CONFIG_MODULES_TREE_LOOKUP case.
> If so then it would make sense IMHO to put the structure definition
> into generic code so that both variants benefit from the shared
> cacheline?
Isn't this optimizing hopeless code? I mean, I can make the change;
something like the below. Although I suppose we should use
____cacheline_aligned here and just take the false sharing.
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -230,11 +230,15 @@ static struct module *mod_find(unsigned
#else /* MODULES_TREE_LOOKUP */
-/*
- * Bounds of module allocation, for speeding up __module_address.
- * Protected by module_mutex.
- */
-static unsigned long module_addr_min = -1UL, module_addr_max = 0;
+static struct mod_bounds {
+ unsigned long addr_min;
+ unsigned long addr_max;
+} mod_bounds __cacheline_aligned = {
+ .addr_min = -1UL,
+};
+
+#define module_addr_min mod_bounds.addr_min
+#define module_addr_max mod_bounds.addr_max
static void mod_tree_insert(struct module *mod) { }
static void mod_tree_remove_init(struct module *mod) { }
@@ -254,6 +258,10 @@ static struct module *mod_find(unsigned
#endif /* MODULES_TREE_LOOKUP */
+/*
+ * Bounds of module text, for speeding up __module_address.
+ * Protected by module_mutex.
+ */
static void __mod_update_bounds(void *base, unsigned int size)
{
unsigned long min = (unsigned long)base;
@@ -3363,8 +3371,8 @@ static int add_unformed_module(struct mo
err = -EEXIST;
goto out;
}
- list_add_rcu(&mod->list, &modules);
mod_update_bounds(mod);
+ list_add_rcu(&mod->list, &modules);
mod_tree_insert(mod);
err = 0;
next prev parent reply other threads:[~2015-04-14 12:57 UTC|newest]
Thread overview: 46+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-13 14:11 [PATCH v5 00/10] latched RB-trees and __module_address() Peter Zijlstra
2015-04-13 14:11 ` [PATCH v5 01/10] module: Sanitize RCU usage and locking Peter Zijlstra
2015-04-13 15:32 ` Ingo Molnar
2015-04-13 15:40 ` Peter Zijlstra
2015-04-13 16:32 ` Ingo Molnar
2015-04-13 14:11 ` [PATCH v5 02/10] module: Annotate module version magic Peter Zijlstra
2015-04-13 14:11 ` [PATCH v5 03/10] module, jump_label: Fix module locking Peter Zijlstra
2015-04-13 14:11 ` [PATCH v5 04/10] rbtree: Make lockless searches non-fatal Peter Zijlstra
2015-04-13 15:50 ` Ingo Molnar
2015-04-13 19:10 ` Peter Zijlstra
2015-04-13 14:11 ` [PATCH v5 05/10] seqlock: Better document raw_write_seqcount_latch() Peter Zijlstra
2015-04-13 16:32 ` Ingo Molnar
2015-04-13 17:08 ` Mathieu Desnoyers
2015-04-13 17:43 ` Paul E. McKenney
2015-04-13 18:21 ` Linus Torvalds
2015-04-13 18:42 ` Paul E. McKenney
2015-04-14 10:25 ` Ingo Molnar
2015-04-14 13:04 ` Paul E. McKenney
2015-04-14 14:31 ` Ingo Molnar
2015-04-14 15:11 ` Paul E. McKenney
2015-04-13 19:17 ` Peter Zijlstra
2015-04-13 19:47 ` Peter Zijlstra
2015-04-14 10:26 ` Ingo Molnar
2015-04-13 14:11 ` [PATCH v5 06/10] rbtree: Implement generic latch_tree Peter Zijlstra
2015-04-13 16:43 ` Ingo Molnar
2015-04-13 19:20 ` Peter Zijlstra
2015-04-13 14:11 ` [PATCH v5 07/10] module: Optimize __module_address() using a latched RB-tree Peter Zijlstra
2015-04-13 16:49 ` Ingo Molnar
2015-04-14 12:31 ` Peter Zijlstra
2015-04-13 14:11 ` [PATCH v5 08/10] module: Make the mod_tree stuff conditional on PERF_EVENTS || TRACING Peter Zijlstra
2015-04-13 16:53 ` Ingo Molnar
2015-04-13 14:11 ` [PATCH v5 09/10] module: Use __module_address() for module_address_lookup() Peter Zijlstra
2015-04-13 14:11 ` [PATCH v5 10/10] module: Rework module_addr_{min,max} Peter Zijlstra
2015-04-13 16:56 ` Ingo Molnar
2015-04-14 2:55 ` Rusty Russell
2015-04-14 6:42 ` Peter Zijlstra
2015-04-14 12:56 ` Peter Zijlstra [this message]
2015-04-14 13:00 ` Ingo Molnar
2015-04-13 17:02 ` [PATCH v5 00/10] latched RB-trees and __module_address() Ingo Molnar
2015-04-14 2:57 ` Rusty Russell
2015-04-14 6:41 ` Peter Zijlstra
2015-04-15 4:41 ` Rusty Russell
2015-04-15 9:41 ` Peter Zijlstra
2015-05-28 2:07 ` Rusty Russell
2015-05-28 11:20 ` Peter Zijlstra
2015-05-28 23:49 ` Rusty Russell
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=20150414125657.GM5029@twins.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=andi@firstfloor.org \
--cc=laijs@cn.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@horizon.com \
--cc=mathieu.desnoyers@efficios.com \
--cc=mingo@kernel.org \
--cc=oleg@redhat.com \
--cc=paulmck@linux.vnet.ibm.com \
--cc=rostedt@goodmis.org \
--cc=rusty@rustcorp.com.au \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/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