From: Frank Rowand <frowand.list@gmail.com>
To: Rob Herring <robh+dt@kernel.org>, cpandya@codeaurora.org
Cc: devicetree@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] of: cache phandle nodes to decrease cost of of_find_node_by_phandle()
Date: Wed, 31 Jan 2018 12:07:32 -0800 [thread overview]
Message-ID: <4d849dcd-5186-fe60-97dd-8481846aac4e@gmail.com> (raw)
In-Reply-To: <1517428977-25653-1-git-send-email-frowand.list@gmail.com>
Hi Rob,
Please ignore this one. My signed-off is below the first "---".
-Frank
On 01/31/18 12:02, frowand.list@gmail.com wrote:
> From: Frank Rowand <frank.rowand@sony.com>
>
> Create a cache of the nodes that contain a phandle property. Use this
> cache to find the node for a given phandle value instead of scanning
> the devicetree to find the node. If the phandle value is not found
> in the cache, of_find_node_by_phandle() will fall back to the tree
> scan algorithm.
>
> The cache is initialized in of_core_init().
>
> The cache is freed via a late_initcall_sync().
>
> ---
>
> Some of_find_by_phandle() calls may occur before the cache is
> initialized or after it is freed. For example, for the qualcomm
> qcom-apq8074-dragonboard, 11 calls occur before the initialization
> and 80 occur after the cache is freed (out of 516 total calls.)
>
> Signed-off-by: Frank Rowand <frank.rowand@sony.com>
> ---
> drivers/of/base.c | 85 ++++++++++++++++++++++++++++++++++++++++++++++---
> drivers/of/of_private.h | 5 +++
> drivers/of/resolver.c | 21 ------------
> 3 files changed, 86 insertions(+), 25 deletions(-)
>
> diff --git a/drivers/of/base.c b/drivers/of/base.c
> index 26618ba8f92a..c3091d0e391f 100644
> --- a/drivers/of/base.c
> +++ b/drivers/of/base.c
> @@ -95,10 +95,14 @@ int __weak of_node_to_nid(struct device_node *np)
> }
> #endif
>
> +static void of_populate_phandle_cache(void);
> +
> void __init of_core_init(void)
> {
> struct device_node *np;
>
> + of_populate_phandle_cache();
> +
> /* Create the kset, and register existing nodes */
> mutex_lock(&of_mutex);
> of_kset = kset_create_and_add("devicetree", NULL, firmware_kobj);
> @@ -990,6 +994,72 @@ int of_modalias_node(struct device_node *node, char *modalias, int len)
> }
> EXPORT_SYMBOL_GPL(of_modalias_node);
>
> +phandle live_tree_max_phandle(void)
> +{
> + struct device_node *node;
> + phandle max_phandle;
> + unsigned long flags;
> +
> + raw_spin_lock_irqsave(&devtree_lock, flags);
> + max_phandle = 0;
> + for_each_of_allnodes(node) {
> + if (node->phandle != OF_PHANDLE_ILLEGAL &&
> + node->phandle > max_phandle)
> + max_phandle = node->phandle;
> + }
> + raw_spin_unlock_irqrestore(&devtree_lock, flags);
> +
> + return max_phandle;
> +}
> +
> +static struct device_node **phandle_cache;
> +static u32 max_phandle_cache;
> +
> +static int __init of_free_phandle_cache(void)
> +{
> + max_phandle_cache = 0;
> + kfree(phandle_cache);
> + phandle_cache = NULL;
> +
> + return 0;
> +}
> +late_initcall_sync(of_free_phandle_cache);
> +
> +static void of_populate_phandle_cache(void)
> +{
> + unsigned long flags;
> + phandle max_phandle;
> + u32 nodes = 0;
> + struct device_node *np;
> +
> + if (phandle_cache)
> + return;
> +
> + max_phandle = live_tree_max_phandle();
> +
> + raw_spin_lock_irqsave(&devtree_lock, flags);
> +
> + for_each_of_allnodes(np)
> + nodes++;
> +
> + /* sanity cap for malformed tree */
> + if (max_phandle > nodes)
> + max_phandle = nodes;
> +
> + phandle_cache = kzalloc((max_phandle + 1) * sizeof(*phandle_cache),
> + GFP_KERNEL);
> +
> + for_each_of_allnodes(np)
> + if (np->phandle != OF_PHANDLE_ILLEGAL &&
> + np->phandle <= max_phandle &&
> + np->phandle)
> + phandle_cache[np->phandle] = np;
> +
> + max_phandle_cache = max_phandle;
> +
> + raw_spin_unlock_irqrestore(&devtree_lock, flags);
> +}
> +
> /**
> * of_find_node_by_phandle - Find a node given a phandle
> * @handle: phandle of the node to find
> @@ -999,16 +1069,23 @@ int of_modalias_node(struct device_node *node, char *modalias, int len)
> */
> struct device_node *of_find_node_by_phandle(phandle handle)
> {
> - struct device_node *np;
> + struct device_node *np = NULL;
> unsigned long flags;
>
> if (!handle)
> return NULL;
>
> raw_spin_lock_irqsave(&devtree_lock, flags);
> - for_each_of_allnodes(np)
> - if (np->phandle == handle)
> - break;
> +
> + if (handle <= max_phandle_cache)
> + np = phandle_cache[handle];
> +
> + if (!np || np->phandle != handle) {
> + for_each_of_allnodes(np)
> + if (np->phandle == handle)
> + break;
> + }
> +
> of_node_get(np);
> raw_spin_unlock_irqrestore(&devtree_lock, flags);
> return np;
> diff --git a/drivers/of/of_private.h b/drivers/of/of_private.h
> index 92a9a3687446..77005978d60a 100644
> --- a/drivers/of/of_private.h
> +++ b/drivers/of/of_private.h
> @@ -135,6 +135,11 @@ extern void __of_update_property_sysfs(struct device_node *np,
> extern void __of_sysfs_remove_bin_file(struct device_node *np,
> struct property *prop);
>
> +/* illegal phandle value (set when unresolved) */
> +#define OF_PHANDLE_ILLEGAL 0xdeadbeef
> +
> +extern phandle live_tree_max_phandle(void);
> +
> /* iterators for transactions, used for overlays */
> /* forward iterator */
> #define for_each_transaction_entry(_oft, _te) \
> diff --git a/drivers/of/resolver.c b/drivers/of/resolver.c
> index cfaeef5f6cb1..0384ce8fdc3b 100644
> --- a/drivers/of/resolver.c
> +++ b/drivers/of/resolver.c
> @@ -22,27 +22,6 @@
>
> #include "of_private.h"
>
> -/* illegal phandle value (set when unresolved) */
> -#define OF_PHANDLE_ILLEGAL 0xdeadbeef
> -
> -static phandle live_tree_max_phandle(void)
> -{
> - struct device_node *node;
> - phandle phandle;
> - unsigned long flags;
> -
> - raw_spin_lock_irqsave(&devtree_lock, flags);
> - phandle = 0;
> - for_each_of_allnodes(node) {
> - if (node->phandle != OF_PHANDLE_ILLEGAL &&
> - node->phandle > phandle)
> - phandle = node->phandle;
> - }
> - raw_spin_unlock_irqrestore(&devtree_lock, flags);
> -
> - return phandle;
> -}
> -
> static void adjust_overlay_phandles(struct device_node *overlay,
> int phandle_delta)
> {
>
next prev parent reply other threads:[~2018-01-31 20:07 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-01-31 20:02 frowand.list
2018-01-31 20:07 ` Frank Rowand [this message]
2018-02-14 0:25 ` [of] 965aa3fac0: BUG:sleeping_function_called_from_invalid_context_at_mm/slab.h kernel test robot
2018-01-31 20:05 [PATCH] of: cache phandle nodes to decrease cost of of_find_node_by_phandle() frowand.list
2018-01-31 20:10 ` Frank Rowand
2018-01-31 21:43 ` Frank Rowand
2018-02-01 14:24 ` Rob Herring
2018-02-01 14:26 ` Rob Herring
2018-02-01 21:09 ` Frank Rowand
2018-02-02 3:45 ` Rob Herring
2018-02-02 22:34 ` Frank Rowand
2018-02-02 5:53 ` Chintan Pandya
2018-02-03 3:55 ` Frank Rowand
2018-02-01 6:45 ` Chintan Pandya
2018-02-01 8:59 ` Frank Rowand
2018-02-01 10:31 ` Chintan Pandya
2018-02-01 19:10 ` Frank Rowand
2018-02-02 5:59 ` Chintan Pandya
2018-02-02 21:26 ` Frank Rowand
2018-02-05 12:23 ` Chintan Pandya
2018-02-07 12:44 ` Chintan Pandya
2018-02-07 20:09 ` Frank Rowand
2018-02-01 14:34 ` Rob Herring
2018-02-01 21:13 ` Frank Rowand
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=4d849dcd-5186-fe60-97dd-8481846aac4e@gmail.com \
--to=frowand.list@gmail.com \
--cc=cpandya@codeaurora.org \
--cc=devicetree@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=robh+dt@kernel.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