qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Igor Mammedov <imammedo@redhat.com>
To: Tao Xu <tao3.xu@intel.com>
Cc: lvivier@redhat.com, thuth@redhat.com, ehabkost@redhat.com,
	mst@redhat.com, qemu-devel@nongnu.org, jingqi.liu@intel.com,
	fan.du@intel.com, mdroth@linux.vnet.ibm.com, armbru@redhat.com,
	jonathan.cameron@huawei.com
Subject: Re: [PATCH v15 10/12] hmat acpi: Build System Locality Latency and Bandwidth Information Structure(s)
Date: Fri, 8 Nov 2019 15:11:11 +0100	[thread overview]
Message-ID: <20191108151111.726f37df@redhat.com> (raw)
In-Reply-To: <20191107074511.14304-11-tao3.xu@intel.com>

On Thu,  7 Nov 2019 15:45:09 +0800
Tao Xu <tao3.xu@intel.com> wrote:

> From: Liu Jingqi <jingqi.liu@intel.com>
> 
> This structure describes the memory access latency and bandwidth
> information from various memory access initiator proximity domains.
> The latency and bandwidth numbers represented in this structure
> correspond to rated latency and bandwidth for the platform.
> The software could use this information as hint for optimization.
> 
> Signed-off-by: Liu Jingqi <jingqi.liu@intel.com>
> Signed-off-by: Tao Xu <tao3.xu@intel.com>
> ---
> 
> No changes in v15.
> 
> Changes in v13:
>     - Calculate the entries in a new patch.
> ---
>  hw/acpi/hmat.c | 96 +++++++++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 95 insertions(+), 1 deletion(-)
> 
> diff --git a/hw/acpi/hmat.c b/hw/acpi/hmat.c
> index c595098ba7..6ec1310e62 100644
> --- a/hw/acpi/hmat.c
> +++ b/hw/acpi/hmat.c
> @@ -27,6 +27,7 @@
>  #include "qemu/osdep.h"
>  #include "sysemu/numa.h"
>  #include "hw/acpi/hmat.h"
> +#include "qemu/error-report.h"
>  
>  /*
>   * ACPI 6.3:
> @@ -67,11 +68,81 @@ static void build_hmat_mpda(GArray *table_data, uint16_t flags,
>      build_append_int_noprefix(table_data, 0, 8);
>  }
>  
> +/*
> + * ACPI 6.3: 5.2.27.4 System Locality Latency and Bandwidth Information
> + * Structure: Table 5-146
> + */
> +static void build_hmat_lb(GArray *table_data, HMAT_LB_Info *hmat_lb,
> +                          uint32_t num_initiator, uint32_t num_target,
> +                          uint32_t *initiator_list)
> +{
> +    int i;
> +    uint16_t *lb_data;
> +    uint32_t base;
> +    /*
> +     * Length in bytes for entire structure, including 32 bytes of
> +     * fixed length, length of initiator proximity domain list,
> +     * length of target proximity domain list and length of entries
> +     * provides latency/bandwidth values.
> +     */
> +    uint32_t lb_length = 32 + 4 * num_initiator + 4 * num_target +
> +                              2 * num_initiator * num_target;
might be described better using field names from spec, as:

  uint32_t len = 32 /* table length upto and including Entry Base Unit */
                 + 4 * num_initiator /* Initiator Proximity Domain List */
                 + 4 * num_target /* Target Proximity Domain */
                 ....

> +
> +    /* Type */
> +    build_append_int_noprefix(table_data, 1, 2);
> +    /* Reserved */
> +    build_append_int_noprefix(table_data, 0, 2);
> +    /* Length */
> +    build_append_int_noprefix(table_data, lb_length, 4);
> +    /* Flags: Bits [3:0] Memory Hierarchy, Bits[7:4] Reserved */
> +    assert(!(hmat_lb->hierarchy >> 4));
> +    build_append_int_noprefix(table_data, hmat_lb->hierarchy, 1);
> +    /* Data Type */
> +    build_append_int_noprefix(table_data, hmat_lb->data_type, 1);
> +    /* Reserved */
> +    build_append_int_noprefix(table_data, 0, 2);
> +    /* Number of Initiator Proximity Domains (s) */
> +    build_append_int_noprefix(table_data, num_initiator, 4);
> +    /* Number of Target Proximity Domains (t) */
> +    build_append_int_noprefix(table_data, num_target, 4);
> +    /* Reserved */
> +    build_append_int_noprefix(table_data, 0, 4);
> +

> +    if (hmat_lb->data_type <= HMAT_LB_DATA_WRITE_LATENCY) {
> +        base = hmat_lb->base_latency;
> +        lb_data = hmat_lb->entry_latency;
> +    } else {
> +        base = hmat_lb->base_bandwidth;
> +        lb_data = hmat_lb->entry_bandwidth;
> +    }
I'd move this out into caller hmat_build_table_structs() and instead of
hmat_lb pass as arguments data_type, base and data, to keep subtable
API independent of internal NUMA structures.

On this note, considering that a single hmat_lb can represent either
latency or bandwidth only, I'd suggest to unify
   hmat_lb->base_latency/hmat_lb->base_bandwidth
     into hmat_lb->base
   
and do the same for data entries and use hmat_lb->data_type as discriminator.

That should simplify code here and in CLI parsing part as well,
I don't see any reason to keep them separate. 

> +
> +    /* Entry Base Unit */
> +    build_append_int_noprefix(table_data, base, 8);
> +
> +    /* Initiator Proximity Domain List */
> +    for (i = 0; i < num_initiator; i++) {
> +        build_append_int_noprefix(table_data, initiator_list[i], 4);
> +    }
> +
> +    /* Target Proximity Domain List */
> +    for (i = 0; i < num_target; i++) {
> +        build_append_int_noprefix(table_data, i, 4);
> +    }
> +
> +    /* Latency or Bandwidth Entries */
> +    for (i = 0; i < num_initiator * num_target; i++) {
> +        build_append_int_noprefix(table_data, lb_data[i], 2);
> +    }
> +}
> +
>  /* Build HMAT sub table structures */
>  static void hmat_build_table_structs(GArray *table_data, NumaState *numa_state)
>  {
>      uint16_t flags;
> -    int i;
> +    uint32_t num_initiator = 0;
> +    uint32_t initiator_list[MAX_NODES];
> +    int i, hierarchy, type;
> +    HMAT_LB_Info *hmat_lb;
>  
>      for (i = 0; i < numa_state->num_nodes; i++) {
>          flags = 0;
> @@ -82,6 +153,29 @@ static void hmat_build_table_structs(GArray *table_data, NumaState *numa_state)
>  
>          build_hmat_mpda(table_data, flags, numa_state->nodes[i].initiator, i);
>      }
> +
> +    for (i = 0; i < numa_state->num_nodes; i++) {
> +        if (numa_state->nodes[i].has_cpu) {
> +            initiator_list[num_initiator++] = i;
> +        }
> +    }
> +
> +    /*
> +     * ACPI 6.3: 5.2.27.4 System Locality Latency and Bandwidth Information
> +     * Structure: Table 5-146
> +     */
> +    for (hierarchy = HMAT_LB_MEM_MEMORY;
> +         hierarchy <= HMAT_LB_MEM_CACHE_3RD_LEVEL; hierarchy++) {
> +        for (type = HMAT_LB_DATA_ACCESS_LATENCY;
> +             type <= HMAT_LB_DATA_WRITE_BANDWIDTH; type++) {
> +            hmat_lb = numa_state->hmat_lb[hierarchy][type];
> +
> +            if (hmat_lb) {
> +                build_hmat_lb(table_data, hmat_lb, num_initiator,
> +                              numa_state->num_nodes, initiator_list);
> +            }
> +        }
> +    }
>  }
>  
>  void build_hmat(GArray *table_data, BIOSLinker *linker, NumaState *numa_state)



  reply	other threads:[~2019-11-08 14:13 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-11-07  7:44 [PATCH v15 00/12] Build ACPI Heterogeneous Memory Attribute Table (HMAT) Tao Xu
2019-11-07  7:45 ` [PATCH v15 01/12] util/cutils: refactor do_strtosz() to support suffixes list Tao Xu
2019-11-07 17:52   ` Eduardo Habkost
2019-11-07  7:45 ` [PATCH v15 02/12] util/cutils: Add qemu_strtotime_ns() Tao Xu
2019-11-07 17:55   ` Eduardo Habkost
2019-11-07  7:45 ` [PATCH v15 03/12] qapi: Add builtin type time Tao Xu
2019-11-07  7:45 ` [PATCH v15 04/12] tests: Add test for QAPI " Tao Xu
2019-11-07  7:45 ` [PATCH v15 05/12] numa: Extend CLI to provide initiator information for numa nodes Tao Xu
2019-11-07  7:45 ` [PATCH v15 06/12] numa: Extend CLI to provide memory latency and bandwidth information Tao Xu
2019-11-08 11:42   ` Igor Mammedov
2019-11-07  7:45 ` [PATCH v15 07/12] numa: Calculate hmat latency and bandwidth entry list Tao Xu
2019-11-08 12:12   ` Igor Mammedov
2019-11-07  7:45 ` [PATCH v15 08/12] numa: Extend CLI to provide memory side cache information Tao Xu
2019-11-08 13:34   ` Igor Mammedov
2019-11-07  7:45 ` [PATCH v15 09/12] hmat acpi: Build Memory Proximity Domain Attributes Structure(s) Tao Xu
2019-11-08 13:44   ` Igor Mammedov
2019-11-11  1:21     ` Tao Xu
2019-11-07  7:45 ` [PATCH v15 10/12] hmat acpi: Build System Locality Latency and Bandwidth Information Structure(s) Tao Xu
2019-11-08 14:11   ` Igor Mammedov [this message]
2019-11-07  7:45 ` [PATCH v15 11/12] hmat acpi: Build Memory Side Cache " Tao Xu
2019-11-08 14:39   ` Igor Mammedov
2019-11-07  7:45 ` [PATCH v15 12/12] tests/bios-tables-test: add test cases for ACPI HMAT Tao Xu
2019-11-08 14:51   ` Igor Mammedov
2019-11-07  9:03 ` [PATCH v15 00/12] Build ACPI Heterogeneous Memory Attribute Table (HMAT) Michael S. Tsirkin
2019-11-08  0:57   ` Tao Xu

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=20191108151111.726f37df@redhat.com \
    --to=imammedo@redhat.com \
    --cc=armbru@redhat.com \
    --cc=ehabkost@redhat.com \
    --cc=fan.du@intel.com \
    --cc=jingqi.liu@intel.com \
    --cc=jonathan.cameron@huawei.com \
    --cc=lvivier@redhat.com \
    --cc=mdroth@linux.vnet.ibm.com \
    --cc=mst@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=tao3.xu@intel.com \
    --cc=thuth@redhat.com \
    /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;
as well as URLs for NNTP newsgroup(s).