From: Frank Rowand <[email protected]>
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() if modules are not
enabled.
If the devicetree is created by the dtc compiler, with all phandle
property values auto generated, then the size required by the cache
could be 4 * (1 + number of phandles) bytes. This results in an O(1)
node lookup cost for a given phandle value. Due to a concern that the
phandle property values might not be consistent with what is generated
by the dtc compiler, a mask has been added to the cache lookup algorithm.
To maintain the O(1) node lookup cost, the size of the cache has been
increased by rounding the number of entries up to the next power of
two.
The overhead of finding the devicetree node containing a given phandle
value has been noted by several people in the recent past, in some cases
with a patch to add a hashed index of devicetree nodes, based on the
phandle value of the node. One concern with this approach is the extra
space added to each node. This patch takes advantage of the phandle
property values auto generated by the dtc compiler, which begin with
one and monotonically increase by one, resulting in a range of 1..n
for n phandle values. This implementation should also provide a good
reduction of overhead for any range of phandle values that are mostly
in a monotonic range.
Performance measurements by Chintan Pandya <[email protected]>
of several implementations of patches that are similar to this one
suggest an expected reduction of boot time by ~400ms for his test
system. If the cache size was decreased to 64 entries, the boot
time was reduced by ~340 ms. The measurements were on a 4.9.73 kernel
for arch/arm64/boot/dts/qcom/sda670-mtp.dts, contains 2371 nodes and
814 phandle values.
Reported-by: Chintan Pandya <[email protected]>
Signed-off-by: Frank Rowand <[email protected]>
---
A follow on patch will add an early boot allocation of the cache.
Changes since v2:
- add mask to calculation of phandle cache entry
- which results in better overhead reduction for devicetrees with
phandle properties not allocated in the monotonically increasing
range of 1..n
- due to mask, number of entries in cache potentially increased to
next power of two
- minor fixes as suggested by reviewers
- no longer using live_tree_max_phandle() so do not move it from
drivers/of/resolver.c to drivers/of/base.c
Changes since v1:
- change short description from
of: cache phandle nodes to reduce cost of of_find_node_by_phandle()
- rebase on v4.16-rc1
- reorder new functions in base.c to avoid forward declaration
- add locking around kfree(phandle_cache) for memory ordering
- add explicit check for non-null of phandle_cache in
of_find_node_by_phandle(). There is already a check for !handle,
which prevents accessing a null phandle_cache, but that dependency
is not obvious, so this check makes it more apparent.
- do not free phandle_cache if modules are enabled, so that
cached phandles will be available when modules are loaded
drivers/of/base.c | 83 ++++++++++++++++++++++++++++++++++++++++++++++---
drivers/of/of_private.h | 3 ++
drivers/of/resolver.c | 3 --
3 files changed, 82 insertions(+), 7 deletions(-)
diff --git a/drivers/of/base.c b/drivers/of/base.c
index ad28de96e13f..ab545dfa9173 100644
--- a/drivers/of/base.c
+++ b/drivers/of/base.c
@@ -91,10 +91,69 @@ int __weak of_node_to_nid(struct device_node *np)
}
#endif
+static struct device_node **phandle_cache;
+static u32 phandle_cache_mask;
+
+/*
+ * Assumptions behind phandle_cache implementation:
+ * - phandle property values are in a contiguous range of 1..n
+ *
+ * If the assumptions do not hold, then
+ * - the phandle lookup overhead reduction provided by the cache
+ * will likely be less
+ */
+static void of_populate_phandle_cache(void)
+{
+ unsigned long flags;
+ u32 cache_entries;
+ struct device_node *np;
+ u32 phandles = 0;
+
+ raw_spin_lock_irqsave(&devtree_lock, flags);
+
+ kfree(phandle_cache);
+ phandle_cache = NULL;
+
+ for_each_of_allnodes(np)
+ if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
+ phandles++;
+
+ cache_entries = roundup_pow_of_two(phandles);
+ phandle_cache_mask = cache_entries - 1;
+
+ phandle_cache = kcalloc(cache_entries, sizeof(*phandle_cache),
+ GFP_ATOMIC);
+
+ for_each_of_allnodes(np)
+ if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
+ phandle_cache[np->phandle & phandle_cache_mask] = np;
+
+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+}
+
+#ifndef CONFIG_MODULES
+static int __init of_free_phandle_cache(void)
+{
+ unsigned long flags;
+
+ raw_spin_lock_irqsave(&devtree_lock, flags);
+
+ kfree(phandle_cache);
+ phandle_cache = NULL;
+
+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+
+ return 0;
+}
+late_initcall_sync(of_free_phandle_cache);
+#endif
+
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);
@@ -1021,16 +1080,32 @@ 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;
+ phandle masked_handle;
if (!handle)
return NULL;
raw_spin_lock_irqsave(&devtree_lock, flags);
- for_each_of_allnodes(np)
- if (np->phandle == handle)
- break;
+
+ masked_handle = handle & phandle_cache_mask;
+
+ if (phandle_cache) {
+ if (phandle_cache[masked_handle] &&
+ handle == phandle_cache[masked_handle]->phandle)
+ np = phandle_cache[masked_handle];
+ }
+
+ if (!np) {
+ for_each_of_allnodes(np)
+ if (np->phandle == handle) {
+ if (phandle_cache)
+ phandle_cache[masked_handle] = np;
+ 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 0c609e7d0334..fa70650136b4 100644
--- a/drivers/of/of_private.h
+++ b/drivers/of/of_private.h
@@ -131,6 +131,9 @@ 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
+
/* 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 740d19bde601..b2ca8185c8c6 100644
--- a/drivers/of/resolver.c
+++ b/drivers/of/resolver.c
@@ -19,9 +19,6 @@
#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;
--
Frank Rowand <[email protected]>
On 2/15/2018 6:22 AM, [email protected] wrote:
> From: Frank Rowand <[email protected]>
>
> 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() if modules are not
> enabled.
>
> If the devicetree is created by the dtc compiler, with all phandle
> property values auto generated, then the size required by the cache
> could be 4 * (1 + number of phandles) bytes. This results in an O(1)
> node lookup cost for a given phandle value. Due to a concern that the
> phandle property values might not be consistent with what is generated
> by the dtc compiler, a mask has been added to the cache lookup algorithm.
> To maintain the O(1) node lookup cost, the size of the cache has been
> increased by rounding the number of entries up to the next power of
> two.
>
> The overhead of finding the devicetree node containing a given phandle
> value has been noted by several people in the recent past, in some cases
> with a patch to add a hashed index of devicetree nodes, based on the
> phandle value of the node. One concern with this approach is the extra
> space added to each node. This patch takes advantage of the phandle
> property values auto generated by the dtc compiler, which begin with
> one and monotonically increase by one, resulting in a range of 1..n
> for n phandle values. This implementation should also provide a good
> reduction of overhead for any range of phandle values that are mostly
> in a monotonic range.
>
> Performance measurements by Chintan Pandya <[email protected]>
> of several implementations of patches that are similar to this one
> suggest an expected reduction of boot time by ~400ms for his test
> system. If the cache size was decreased to 64 entries, the boot
> time was reduced by ~340 ms. The measurements were on a 4.9.73 kernel
> for arch/arm64/boot/dts/qcom/sda670-mtp.dts, contains 2371 nodes and
> 814 phandle values.
>
> Reported-by: Chintan Pandya <[email protected]>
> Signed-off-by: Frank Rowand <[email protected]>
> ---
>
> A follow on patch will add an early boot allocation of the cache.
>
> Changes since v2:
> - add mask to calculation of phandle cache entry
> - which results in better overhead reduction for devicetrees with
> phandle properties not allocated in the monotonically increasing
> range of 1..n
> - due to mask, number of entries in cache potentially increased to
> next power of two
> - minor fixes as suggested by reviewers
> - no longer using live_tree_max_phandle() so do not move it from
> drivers/of/resolver.c to drivers/of/base.c
>
> Changes since v1:
> - change short description from
> of: cache phandle nodes to reduce cost of of_find_node_by_phandle()
> - rebase on v4.16-rc1
> - reorder new functions in base.c to avoid forward declaration
> - add locking around kfree(phandle_cache) for memory ordering
> - add explicit check for non-null of phandle_cache in
> of_find_node_by_phandle(). There is already a check for !handle,
> which prevents accessing a null phandle_cache, but that dependency
> is not obvious, so this check makes it more apparent.
> - do not free phandle_cache if modules are enabled, so that
> cached phandles will be available when modules are loaded
>
> drivers/of/base.c | 83 ++++++++++++++++++++++++++++++++++++++++++++++---
> drivers/of/of_private.h | 3 ++
> drivers/of/resolver.c | 3 --
> 3 files changed, 82 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/of/base.c b/drivers/of/base.c
> index ad28de96e13f..ab545dfa9173 100644
> --- a/drivers/of/base.c
> +++ b/drivers/of/base.c
> @@ -91,10 +91,69 @@ int __weak of_node_to_nid(struct device_node *np)
> }
> #endif
>
> +static struct device_node **phandle_cache;
> +static u32 phandle_cache_mask;
> +
> +/*
> + * Assumptions behind phandle_cache implementation:
> + * - phandle property values are in a contiguous range of 1..n
> + *
> + * If the assumptions do not hold, then
> + * - the phandle lookup overhead reduction provided by the cache
> + * will likely be less
> + */
> +static void of_populate_phandle_cache(void)
> +{
> + unsigned long flags;
> + u32 cache_entries;
> + struct device_node *np;
> + u32 phandles = 0;
> +
> + raw_spin_lock_irqsave(&devtree_lock, flags);
> +
> + kfree(phandle_cache);
I couldn't understood this. Everything else looks good to me.
> + phandle_cache = NULL;
> +
> + for_each_of_allnodes(np)
> + if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
> + phandles++;
> +
> + cache_entries = roundup_pow_of_two(phandles);
> + phandle_cache_mask = cache_entries - 1;
> +
> + phandle_cache = kcalloc(cache_entries, sizeof(*phandle_cache),
> + GFP_ATOMIC);
> +
> + for_each_of_allnodes(np)
> + if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
> + phandle_cache[np->phandle & phandle_cache_mask] = np;
> +
> + raw_spin_unlock_irqrestore(&devtree_lock, flags);
> +}
> +
> +#ifndef CONFIG_MODULES
> +static int __init of_free_phandle_cache(void)
> +{
> + unsigned long flags;
> +
> + raw_spin_lock_irqsave(&devtree_lock, flags);
> +
> + kfree(phandle_cache);
> + phandle_cache = NULL;
> +
> + raw_spin_unlock_irqrestore(&devtree_lock, flags);
> +
> + return 0;
> +}
> +late_initcall_sync(of_free_phandle_cache);
> +#endif
> +
> 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);
> @@ -1021,16 +1080,32 @@ 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;
> + phandle masked_handle;
>
> if (!handle)
> return NULL;
>
> raw_spin_lock_irqsave(&devtree_lock, flags);
> - for_each_of_allnodes(np)
> - if (np->phandle == handle)
> - break;
> +
> + masked_handle = handle & phandle_cache_mask;
> +
> + if (phandle_cache) {
> + if (phandle_cache[masked_handle] &&
> + handle == phandle_cache[masked_handle]->phandle)
> + np = phandle_cache[masked_handle];
> + }
> +
> + if (!np) {
> + for_each_of_allnodes(np)
> + if (np->phandle == handle) {
> + if (phandle_cache)
> + phandle_cache[masked_handle] = np;
> + 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 0c609e7d0334..fa70650136b4 100644
> --- a/drivers/of/of_private.h
> +++ b/drivers/of/of_private.h
> @@ -131,6 +131,9 @@ 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
> +
> /* 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 740d19bde601..b2ca8185c8c6 100644
> --- a/drivers/of/resolver.c
> +++ b/drivers/of/resolver.c
> @@ -19,9 +19,6 @@
>
> #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;
>
Chintan
--
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center,
Inc. is a member of the Code Aurora Forum, a Linux Foundation
Collaborative Project
On 02/16/18 01:04, Chintan Pandya wrote:
>
>
> On 2/15/2018 6:22 AM, [email protected] wrote:
>> From: Frank Rowand <[email protected]>
>>
>> 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.
>>
< snip >
>> diff --git a/drivers/of/base.c b/drivers/of/base.c
>> index ad28de96e13f..ab545dfa9173 100644
>> --- a/drivers/of/base.c
>> +++ b/drivers/of/base.c
>> @@ -91,10 +91,69 @@ int __weak of_node_to_nid(struct device_node *np)
>> }
>> #endif
>> +static struct device_node **phandle_cache;
>> +static u32 phandle_cache_mask;
>> +
>> +/*
>> + * Assumptions behind phandle_cache implementation:
>> + * - phandle property values are in a contiguous range of 1..n
>> + *
>> + * If the assumptions do not hold, then
>> + * - the phandle lookup overhead reduction provided by the cache
>> + * will likely be less
>> + */
>> +static void of_populate_phandle_cache(void)
>> +{
>> + unsigned long flags;
>> + u32 cache_entries;
>> + struct device_node *np;
>> + u32 phandles = 0;
>> +
>> + raw_spin_lock_irqsave(&devtree_lock, flags);
>> +
>> + kfree(phandle_cache);
>
> I couldn't understood this. Everything else looks good to me.
I will be adding a call to of_populate_phandle_cache() from the
devicetree overlay code. I put the kfree here so that the previous
cache memory is freed when a new cache is created.
Adding the call from the overlay code is not done in this
series because I have a patch series modifying overlays and
I do not want to create a conflict or ordering between that
series and that patch. The lack of the call from overlay
code means that overlay code will gain some of the overhead
reduction from this patch, but possibly not the entire reduction.
>
>> + phandle_cache = NULL;
>> +
>> + for_each_of_allnodes(np)
< snip >
On 02/16/18 14:20, Frank Rowand wrote:
> On 02/16/18 01:04, Chintan Pandya wrote:
>>
>>
>> On 2/15/2018 6:22 AM, [email protected] wrote:
>>> From: Frank Rowand <[email protected]>
>>>
>>> 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.
>>>
>
> < snip >
>
>>> diff --git a/drivers/of/base.c b/drivers/of/base.c
>>> index ad28de96e13f..ab545dfa9173 100644
>>> --- a/drivers/of/base.c
>>> +++ b/drivers/of/base.c
>>> @@ -91,10 +91,69 @@ int __weak of_node_to_nid(struct device_node *np)
>>> }
>>> #endif
>>> +static struct device_node **phandle_cache;
>>> +static u32 phandle_cache_mask;
>>> +
>>> +/*
>>> + * Assumptions behind phandle_cache implementation:
>>> + * - phandle property values are in a contiguous range of 1..n
>>> + *
>>> + * If the assumptions do not hold, then
>>> + * - the phandle lookup overhead reduction provided by the cache
>>> + * will likely be less
>>> + */
>>> +static void of_populate_phandle_cache(void)
>>> +{
>>> + unsigned long flags;
>>> + u32 cache_entries;
>>> + struct device_node *np;
>>> + u32 phandles = 0;
>>> +
>>> + raw_spin_lock_irqsave(&devtree_lock, flags);
>>> +
>>> + kfree(phandle_cache);
>>
>> I couldn't understood this. Everything else looks good to me.
>
> I will be adding a call to of_populate_phandle_cache() from the
> devicetree overlay code. I put the kfree here so that the previous
> cache memory is freed when a new cache is created.
>
> Adding the call from the overlay code is not done in this
> series because I have a patch series modifying overlays and
> I do not want to create a conflict or ordering between that
> series and that patch. The lack of the call from overlay
^^^^ this
> code means that overlay code will gain some of the overhead
> reduction from this patch, but possibly not the entire reduction.
>
>
>>
>>> + phandle_cache = NULL;
>>> +
>>> + for_each_of_allnodes(np)
>
> < snip >
>
On 2/15/2018 6:22 AM, [email protected] wrote:
> +static void of_populate_phandle_cache(void)
> +{
> + unsigned long flags;
> + u32 cache_entries;
> + struct device_node *np;
> + u32 phandles = 0;
> +
> + raw_spin_lock_irqsave(&devtree_lock, flags);
> +
> + kfree(phandle_cache);
> + phandle_cache = NULL;
> +
> + for_each_of_allnodes(np)
> + if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
> + phandles++;
> +
> + cache_entries = roundup_pow_of_two(phandles);
> + phandle_cache_mask = cache_entries - 1;
> +
> + phandle_cache = kcalloc(cache_entries, sizeof(*phandle_cache),
> + GFP_ATOMIC);
> +
NULL check of phandle_cache would be good to have.
> + for_each_of_allnodes(np)
> + if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
> + phandle_cache[np->phandle & phandle_cache_mask] = np;
> +
> + raw_spin_unlock_irqrestore(&devtree_lock, flags);
> +}
Chintan
--
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center,
Inc. is a member of the Code Aurora Forum, a Linux Foundation
Collaborative Project
On 02/28/18 05:27, Chintan Pandya wrote:
>
>
> On 2/15/2018 6:22 AM, [email protected] wrote:
>
>> +static void of_populate_phandle_cache(void)
>> +{
>> + unsigned long flags;
>> + u32 cache_entries;
>> + struct device_node *np;
>> + u32 phandles = 0;
>> +
>> + raw_spin_lock_irqsave(&devtree_lock, flags);
>> +
>> + kfree(phandle_cache);
>> + phandle_cache = NULL;
>> +
>> + for_each_of_allnodes(np)
>> + if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
>> + phandles++;
>> +
>> + cache_entries = roundup_pow_of_two(phandles);
>> + phandle_cache_mask = cache_entries - 1;
>> +
>> + phandle_cache = kcalloc(cache_entries, sizeof(*phandle_cache),
>> + GFP_ATOMIC);
>> +
>
> NULL check of phandle_cache would be good to have.
Thanks for catching that.
>> + for_each_of_allnodes(np)
>> + if (np->phandle && np->phandle != OF_PHANDLE_ILLEGAL)
>> + phandle_cache[np->phandle & phandle_cache_mask] = np;
>> +
>> + raw_spin_unlock_irqrestore(&devtree_lock, flags);
>> +}
>
>
> Chintan