Message-ID: <4ee4af4d-d987-33f9-4d55-67bfede3ebfa@caviumnetworks.com>
Date:   Wed, 18 Oct 2017 07:39:25 +0200
From:   Tomasz Nowicki <tnowicki@...iumnetworks.com>
To:     Jeremy Linton <jeremy.linton@....com>, linux-acpi@...r.kernel.org
Cc:     mark.rutland@....com, Jonathan.Zhang@...ium.com,
        Jayachandran.Nair@...ium.com, lorenzo.pieralisi@....com,
        catalin.marinas@....com, gregkh@...uxfoundation.org,
        jhugo@...eaurora.org, rjw@...ysocki.net, linux-pm@...r.kernel.org,
        will.deacon@....com, linux-kernel@...r.kernel.org, ahs3@...hat.com,
        viresh.kumar@...aro.org, hanjun.guo@...aro.org,
        sudeep.holla@....com, austinwc@...eaurora.org,
        wangxiongfeng2@...wei.com, linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH v3 1/7] ACPI/PPTT: Add Processor Properties Topology Table
 parsing

Hi,

On 17.10.2017 17:22, Jeremy Linton wrote:
> Hi,
> 
> On 10/17/2017 08:25 AM, Tomasz Nowicki wrote:
>> Hi Jeremy,
>>
>> I did second round of review and have some more comments, please see 
>> below:
>>
>> On 12.10.2017 21:48, Jeremy Linton wrote:
>>> ACPI 6.2 adds a new table, which describes how processing units
>>> are related to each other in tree like fashion. Caches are
>>> also sprinkled throughout the tree and describe the properties
>>> of the caches in relation to other caches and processing units.
>>>
>>> Add the code to parse the cache hierarchy and report the total
>>> number of levels of cache for a given core using
>>> acpi_find_last_cache_level() as well as fill out the individual
>>> cores cache information with cache_setup_acpi() once the
>>> cpu_cacheinfo structure has been populated by the arch specific
>>> code.
>>>
>>> Further, report peers in the topology using setup_acpi_cpu_topology()
>>> to report a unique ID for each processing unit at a given level
>>> in the tree. These unique id's can then be used to match related
>>> processing units which exist as threads, COD (clusters
>>> on die), within a given package, etc.
>>>
>>> Signed-off-by: Jeremy Linton <jeremy.linton@....com>
>>> ---
>>>   drivers/acpi/pptt.c | 485 
>>> ++++++++++++++++++++++++++++++++++++++++++++++++++++
>>>   1 file changed, 485 insertions(+)
>>>   create mode 100644 drivers/acpi/pptt.c
>>>
>>> diff --git a/drivers/acpi/pptt.c b/drivers/acpi/pptt.c
>>> new file mode 100644
>>> index 000000000000..c86715fed4a7
>>> --- /dev/null
>>> +++ b/drivers/acpi/pptt.c
>>> @@ -0,1 +1,485 @@
>>> +/*
>>> + * Copyright (C) 2017, ARM
>>> + *
>>> + * This program is free software; you can redistribute it and/or 
>>> modify it
>>> + * under the terms and conditions of the GNU General Public License,
>>> + * version 2, as published by the Free Software Foundation.
>>> + *
>>> + * This program is distributed in the hope it will be useful, but 
>>> WITHOUT
>>> + * ANY WARRANTY; without even the implied warranty of 
>>> MERCHANTABILITY or
>>> + * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public 
>>> License for
>>> + * more details.
>>> + *
>>> + * This file implements parsing of Processor Properties Topology 
>>> Table (PPTT)
>>> + * which is optionally used to describe the processor and cache 
>>> topology.
>>> + * Due to the relative pointers used throughout the table, this doesn't
>>> + * leverage the existing subtable parsing in the kernel.
>>> + */
>>> +#define pr_fmt(fmt) "ACPI PPTT: " fmt
>>> +
>>> +#include <linux/acpi.h>
>>> +#include <linux/cacheinfo.h>
>>> +#include <acpi/processor.h>
>>> +
>>> +/*
>>> + * Given the PPTT table, find and verify that the subtable entry
>>> + * is located within the table
>>> + */
>>> +static struct acpi_subtable_header *fetch_pptt_subtable(
>>> +    struct acpi_table_header *table_hdr, u32 pptt_ref)
>>> +{
>>> +    struct acpi_subtable_header *entry;
>>> +
>>> +    /* there isn't a subtable at reference 0 */
>>> +    if (!pptt_ref)
>>> +        return NULL;
>>> +
>>> +    if (pptt_ref + sizeof(struct acpi_subtable_header) > 
>>> table_hdr->length)
>>> +        return NULL;
>>> +
>>> +    entry = (struct acpi_subtable_header *)((u8 *)table_hdr + 
>>> pptt_ref);
>>> +
>>> +    if (pptt_ref + entry->length > table_hdr->length)
>>> +        return NULL;
>>> +
>>> +    return entry;
>>> +}
>>> +
>>> +static struct acpi_pptt_processor *fetch_pptt_node(
>>> +    struct acpi_table_header *table_hdr, u32 pptt_ref)
>>> +{
>>> +    return (struct acpi_pptt_processor 
>>> *)fetch_pptt_subtable(table_hdr, pptt_ref);
>>> +}
>>> +
>>> +static struct acpi_pptt_cache *fetch_pptt_cache(
>>> +    struct acpi_table_header *table_hdr, u32 pptt_ref)
>>> +{
>>> +    return (struct acpi_pptt_cache *)fetch_pptt_subtable(table_hdr, 
>>> pptt_ref);
>>> +}
>>> +
>>> +static struct acpi_subtable_header *acpi_get_pptt_resource(
>>> +    struct acpi_table_header *table_hdr,
>>> +    struct acpi_pptt_processor *node, int resource)
>>> +{
>>> +    u32 ref;
>>> +
>>> +    if (resource >= node->number_of_priv_resources)
>>> +        return NULL;
>>> +
>>> +    ref = *(u32 *)((u8 *)node + sizeof(struct acpi_pptt_processor) +
>>> +              sizeof(u32) * resource);
>>> +
>>> +    return fetch_pptt_subtable(table_hdr, ref);
>>> +}
>>> +
>>> +/*
>>> + * given a pptt resource, verify that it is a cache node, then walk
>>> + * down each level of caches, counting how many levels are found
>>> + * as well as checking the cache type (icache, dcache, unified). If a
>>> + * level & type match, then we set found, and continue the search.
>>> + * Once the entire cache branch has been walked return its max
>>> + * depth.
>>> + */
>>> +static int acpi_pptt_walk_cache(struct acpi_table_header *table_hdr,
>>> +                int local_level,
>>> +                struct acpi_subtable_header *res,
>>> +                struct acpi_pptt_cache **found,
>>> +                int level, int type)
>>> +{
>>> +    struct acpi_pptt_cache *cache;
>>> +
>>> +    if (res->type != ACPI_PPTT_TYPE_CACHE)
>>> +        return 0;
>>> +
>>> +    cache = (struct acpi_pptt_cache *) res;
>>> +    while (cache) {
>>> +        local_level++;
>>> +
>>> +        if ((local_level == level) &&
>>> +            (cache->flags & ACPI_PPTT_CACHE_TYPE_VALID) &&
>>> +            ((cache->attributes & ACPI_PPTT_MASK_CACHE_TYPE) == 
>>> type)) {
>>
>> Attributes have to be shifted:
>>
>> (cache->attributes & ACPI_PPTT_MASK_CACHE_TYPE) >> 2
> 
> Hmmm, I'm not sure that is true, the top level function in this routine 
> convert the "linux" constant to the ACPI version of that constant. In 
> that case the "type" field is pre-shifted, so that it matches the result 
> of just anding against the field... That is unless I messed something 
> up, which I don't see at the moment (and the code of course has been 
> tested with PPTT's from multiple people at this point).

For ThunderX2 I got lots of errors in dmesg:
Found duplicate cache level/type unable to determine uniqueness

So I fixed "type" macros definitions (without shifting) and shift it 
here which fixes the issue. As you said, it can be pre-shifted as well.

> 
> 
>>
>>> +            if (*found != NULL)
>>> +                pr_err("Found duplicate cache level/type unable to 
>>> determine uniqueness\n");
>>> +
>>> +            pr_debug("Found cache @ level %d\n", level);
>>> +            *found = cache;
>>> +            /*
>>> +             * continue looking at this node's resource list
>>> +             * to verify that we don't find a duplicate
>>> +             * cache node.
>>> +             */
>>> +        }
>>> +        cache = fetch_pptt_cache(table_hdr, 
>>> cache->next_level_of_cache);
>>> +    }
>>> +    return local_level;
>>> +}
>>> +
>>> +/*
>>> + * Given a CPU node look for cache levels that exist at this level, 
>>> and then
>>> + * for each cache node, count how many levels exist below (logically 
>>> above) it.
>>> + * If a level and type are specified, and we find that level/type, 
>>> abort
>>> + * processing and return the acpi_pptt_cache structure.
>>> + */
>>> +static struct acpi_pptt_cache *acpi_find_cache_level(
>>> +    struct acpi_table_header *table_hdr,
>>> +    struct acpi_pptt_processor *cpu_node,
>>> +    int *starting_level, int level, int type)
>>> +{
>>> +    struct acpi_subtable_header *res;
>>> +    int number_of_levels = *starting_level;
>>> +    int resource = 0;
>>> +    struct acpi_pptt_cache *ret = NULL;
>>> +    int local_level;
>>> +
>>> +    /* walk down from the processor node */
>>> +    while ((res = acpi_get_pptt_resource(table_hdr, cpu_node, 
>>> resource))) {
>>> +        resource++;
>>> +
>>> +        local_level = acpi_pptt_walk_cache(table_hdr, *starting_level,
>>> +                           res, &ret, level, type);
>>> +        /*
>>> +         * we are looking for the max depth. Since its potentially
>>> +         * possible for a given node to have resources with differing
>>> +         * depths verify that the depth we have found is the largest.
>>> +         */
>>> +        if (number_of_levels < local_level)
>>> +            number_of_levels = local_level;
>>> +    }
>>> +    if (number_of_levels > *starting_level)
>>> +        *starting_level = number_of_levels;
>>> +
>>> +    return ret;
>>> +}
>>> +
>>> +/*
>>> + * given a processor node containing a processing unit, walk into it 
>>> and count
>>> + * how many levels exist solely for it, and then walk up each level 
>>> until we hit
>>> + * the root node (ignore the package level because it may be 
>>> possible to have
>>> + * caches that exist across packages). Count the number of cache 
>>> levels that
>>> + * exist at each level on the way up.
>>> + */
>>> +static int acpi_process_node(struct acpi_table_header *table_hdr,
>>> +                 struct acpi_pptt_processor *cpu_node)
>>> +{
>>> +    int total_levels = 0;
>>> +
>>> +    do {
>>> +        acpi_find_cache_level(table_hdr, cpu_node, &total_levels, 0, 
>>> 0);
>>> +        cpu_node = fetch_pptt_node(table_hdr, cpu_node->parent);
>>> +    } while (cpu_node);
>>> +
>>> +    return total_levels;
>>> +}
>>> +
>>> +/* determine if the given node is a leaf node */
>>> +static int acpi_pptt_leaf_node(struct acpi_table_header *table_hdr,
>>> +                   struct acpi_pptt_processor *node)
>>> +{
>>> +    struct acpi_subtable_header *entry;
>>> +    unsigned long table_end;
>>> +    u32 node_entry;
>>> +    struct acpi_pptt_processor *cpu_node;
>>> +
>>> +    table_end = (unsigned long)table_hdr + table_hdr->length;
>>> +    node_entry = (u32)((u8 *)node - (u8 *)table_hdr);
>>> +    entry = (struct acpi_subtable_header *)((u8 *)table_hdr +
>>> +                        sizeof(struct acpi_table_pptt));
>>> +
>>> +    while (((unsigned long)entry) + sizeof(struct 
>>> acpi_subtable_header) < table_end) {
>>> +        cpu_node = (struct acpi_pptt_processor *)entry;
>>> +        if ((entry->type == ACPI_PPTT_TYPE_PROCESSOR) &&
>>> +            (cpu_node->parent == node_entry))
>>> +            return 0;
>>> +        entry = (struct acpi_subtable_header *)((u8 *)entry + 
>>> entry->length);
>>> +    }
>>> +    return 1;
>>> +}
>>> +
>>> +/*
>>> + * Find the subtable entry describing the provided processor
>>> + */
>>> +static struct acpi_pptt_processor *acpi_find_processor_node(
>>> +    struct acpi_table_header *table_hdr,
>>> +    u32 acpi_cpu_id)
>>> +{
>>> +    struct acpi_subtable_header *entry;
>>> +    unsigned long table_end;
>>> +    struct acpi_pptt_processor *cpu_node;
>>> +
>>> +    table_end = (unsigned long)table_hdr + table_hdr->length;
>>> +    entry = (struct acpi_subtable_header *)((u8 *)table_hdr +
>>> +                        sizeof(struct acpi_table_pptt));
>>> +
>>> +    /* find the processor structure associated with this cpuid */
>>> +    while (((unsigned long)entry) + sizeof(struct 
>>> acpi_subtable_header) < table_end) {
>>> +        cpu_node = (struct acpi_pptt_processor *)entry;
>>> +
>>> +        if ((entry->type == ACPI_PPTT_TYPE_PROCESSOR) &&
>>> +            acpi_pptt_leaf_node(table_hdr, cpu_node)) {
>>> +            pr_debug("checking phy_cpu_id %d against acpi id %d\n",
>>> +                 acpi_cpu_id, cpu_node->acpi_processor_id);
>>> +            if (acpi_cpu_id == cpu_node->acpi_processor_id) {
>>> +                /* found the correct entry */
>>> +                pr_debug("match found!\n");
>>> +                return (struct acpi_pptt_processor *)entry;
>>> +            }
>>> +        }
>>> +
>>> +        if (entry->length == 0) {
>>> +            pr_err("Invalid zero length subtable\n");
>>> +            break;
>>> +        }
>>> +        entry = (struct acpi_subtable_header *)
>>> +            ((u8 *)entry + entry->length);
>>> +    }
>>> +
>>> +    return NULL;
>>> +}
>>> +
>>> +/*
>>> + * Given a acpi_pptt_processor node, walk up until we identify the
>>> + * package that the node is associated with or we run out of levels
>>> + * to request.
>>> + */
>>> +static struct acpi_pptt_processor *acpi_find_processor_package_id(
>>> +    struct acpi_table_header *table_hdr,
>>> +    struct acpi_pptt_processor *cpu,
>>> +    int level)
>>> +{
>>> +    struct acpi_pptt_processor *prev_node;
>>> +
>>> +    while (cpu && level && !(cpu->flags & 
>>> ACPI_PPTT_PHYSICAL_PACKAGE)) {
>>> +        pr_debug("level %d\n", level);
>>> +        prev_node = fetch_pptt_node(table_hdr, cpu->parent);
>>> +        if (prev_node == NULL)
>>> +            break;
>>> +        cpu = prev_node;
>>> +        level--;
>>> +    }
>>> +    return cpu;
>>> +}
>>> +
>>> +static int acpi_parse_pptt(struct acpi_table_header *table_hdr, u32 
>>> acpi_cpu_id)
>>
>> The function name can be more descriptive. How about:
>>
>> acpi_count_cache_level() ?
> 
> The naming has drifted a bit, so yes, that routine is only used by the 
> portion which is determining the number of cache levels for a given PE.
> 
> 
>>
>>> +{
>>> +    int number_of_levels = 0;
>>> +    struct acpi_pptt_processor *cpu;
>>> +
>>> +    cpu = acpi_find_processor_node(table_hdr, acpi_cpu_id);
>>> +    if (cpu)
>>> +        number_of_levels = acpi_process_node(table_hdr, cpu);
>>> +
>>> +    return number_of_levels;
>>> +}
>>
>> It is hard to follow what acpi_find_cache_level() and 
>> acpi_pptt_walk_cache() really do. It is because they are trying to do 
>> too many things at the same time. IMO, splitting 
>> acpi_find_cache_level() logic to:
>> 1. counting the cache levels (max depth)
>> 2. finding the specific cache node
>> makes sense.
> 
> I disagree, that routine is shared by the two code paths because its 
> functionality is 99% duplicated between the two. The difference being 
> whether it terminates the search at a given level, or continues 
> searching until it runs out of nodes. The latter case is simply a 
> degenerate version of the first.

Mostly it is about trade-off between code simplicity and redundancy, I 
personally prefer the former. It is not the critical issue though.

> 
> 
>>
>> Also, seems like we can merge acpi_parse_pptt() & acpi_process_node().
> 
> That is true, but I fail to see how any of this is actually fixes 
> anything. There are a million ways to do this, including as pointed out 
> by building another data-structure to simplify the parsing what is a 
> table that is less than ideal for runtime parsing (starting with the 
> direction of the relative pointers, and ending with having to "infer" 
> information that isn't directly flagged). I actually built a couple 
> other versions of this, including a nice cute version which is about 1/8 
> this size of this and really easy to understand but of course is 
> recursive...

I believe this will improve code readability. Obviously, you can 
disagree with my suggestions.

Thanks,
Tomasz

Powered by blists - more mailing lists