All of lore.kernel.org
 help / color / mirror / Atom feed
From: fan <nifan.cxl@gmail.com>
To: shiju.jose@huawei.com
Cc: qemu-devel@nongnu.org, linux-cxl@vger.kernel.org,
	jonathan.cameron@huawei.com, tanxiaofei@huawei.com,
	prime.zeng@hisilicon.com, linuxarm@huawei.com,
	fan.ni@samsung.com
Subject: Re: [PATCH v2 2/3] hw/cxl/cxl-mailbox-utils: Add device patrol scrub control feature
Date: Thu, 15 Feb 2024 10:47:16 -0800	[thread overview]
Message-ID: <Zc5cNLH9FS24qXTe@debian> (raw)
In-Reply-To: <20231124135338.1191-3-shiju.jose@huawei.com>

On Fri, Nov 24, 2023 at 09:53:36PM +0800, shiju.jose@huawei.com wrote:
> From: Shiju Jose <shiju.jose@huawei.com>
> 
> CXL spec 3.1 section 8.2.9.9.11.1 describes the device patrol scrub control
> feature. The device patrol scrub proactively locates and makes corrections
> to errors in regular cycle. The patrol scrub control allows the request to
> configure patrol scrub input configurations.
> 
> The patrol scrub control allows the requester to specify the number of
> hours for which the patrol scrub cycles must be completed, provided that
> the requested number is not less than the minimum number of hours for the
> patrol scrub cycle that the device is capable of. In addition, the patrol
> scrub controls allow the host to disable and enable the feature in case
> disabling of the feature is needed for other purposes such as
> performance-aware operations which require the background operations to be
> turned off.
> 
> Reviewed-by: Davidlohr Bueso <dave@stgolabs.net>
> Signed-off-by: Shiju Jose <shiju.jose@huawei.com>
> ---

LGTM except for some minor comments inlined.


>  hw/cxl/cxl-mailbox-utils.c | 97 +++++++++++++++++++++++++++++++++++++-
>  1 file changed, 96 insertions(+), 1 deletion(-)
> 
> diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
> index 1bbc9a48a6..5a6f4e4029 100644
> --- a/hw/cxl/cxl-mailbox-utils.c
> +++ b/hw/cxl/cxl-mailbox-utils.c
> @@ -809,6 +809,7 @@ typedef struct CXLSupportedFeatureEntry {
>  } QEMU_PACKED CXLSupportedFeatureEntry;
>  
>  enum CXL_SUPPORTED_FEATURES_LIST {
> +    CXL_FEATURE_PATROL_SCRUB = 0,
>      CXL_FEATURE_MAX
>  };
>  
> @@ -849,6 +850,37 @@ enum CXL_SET_FEATURE_FLAG_DATA_TRANSFER {
>      CXL_SET_FEATURE_FLAG_DATA_TRANSFER_MAX
>  };
>  
> +/* CXL r3.1 section 8.2.9.9.11.1: Device Patrol Scrub Control Feature */
> +static const QemuUUID patrol_scrub_uuid = {
> +    .data = UUID(0x96dad7d6, 0xfde8, 0x482b, 0xa7, 0x33,
> +                 0x75, 0x77, 0x4e, 0x06, 0xdb, 0x8a)
> +};
> +
> +#define CXL_MEMDEV_PS_GET_FEATURE_VERSION    0x01
> +#define CXL_MEMDEV_PS_SET_FEATURE_VERSION    0x01
> +#define CXL_MEMDEV_PS_SCRUB_CYCLE_CHANGE_CAP_DEFAULT    BIT(0)
> +#define CXL_MEMDEV_PS_SCRUB_REALTIME_REPORT_CAP_DEFAULT    BIT(1)
> +#define CXL_MEMDEV_PS_CUR_SCRUB_CYCLE_DEFAULT    12
> +#define CXL_MEMDEV_PS_MIN_SCRUB_CYCLE_DEFAULT    1
> +#define CXL_MEMDEV_PS_ENABLE_DEFAULT    0
> +
> +/* CXL memdev patrol scrub control attributes */
> +struct CXLMemPatrolScrubReadAttrbs {
> +        uint8_t scrub_cycle_cap;
> +        uint16_t scrub_cycle;
> +        uint8_t scrub_flags;
> +} QEMU_PACKED cxl_memdev_ps_feat_read_attrbs;
> +
> +typedef struct CXLMemPatrolScrubWriteAttrbs {
> +    uint8_t scrub_cycle_hr;
> +    uint8_t scrub_flags;
> +} QEMU_PACKED CXLMemPatrolScrubWriteAttrbs;
> +
> +typedef struct CXLMemPatrolScrubSetFeature {
> +        CXLSetFeatureInHeader hdr;
> +        CXLMemPatrolScrubWriteAttrbs feat_data;
> +} QEMU_PACKED QEMU_ALIGNED(16) CXLMemPatrolScrubSetFeature;
> +
>  /* CXL r3.0 section 8.2.9.6.1: Get Supported Features (Opcode 0500h) */
>  static CXLRetCode cmd_features_get_supported(const struct cxl_cmd *cmd,
>                                               uint8_t *payload_in,
> @@ -872,7 +904,7 @@ static CXLRetCode cmd_features_get_supported(const struct cxl_cmd *cmd,
>      uint16_t feat_entries = 0;
>  
>      if (get_feats_in->count < sizeof(CXLSupportedFeatureHeader) ||
> -        get_feats_in->start_index > CXL_FEATURE_MAX) {
> +        get_feats_in->start_index >= CXL_FEATURE_MAX) {

Not totally sure about this, the spec says "...Greater than..." although I
also think it should be >=. Similar things for the offset usage below.

Fan

>          return CXL_MBOX_INVALID_INPUT;
>      }
>      req_entries = (get_feats_in->count -
> @@ -884,6 +916,31 @@ static CXLRetCode cmd_features_get_supported(const struct cxl_cmd *cmd,
>      entry = 0;
>      while (entry < req_entries) {
>          switch (index) {
> +        case  CXL_FEATURE_PATROL_SCRUB:
> +            /* Fill supported feature entry for device patrol scrub control */
> +            get_feats_out->feat_entries[entry] =
> +                           (struct CXLSupportedFeatureEntry) {
> +                .uuid = patrol_scrub_uuid,
> +                .feat_index = index,
> +                .get_feat_size = sizeof(cxl_memdev_ps_feat_read_attrbs),
> +                .set_feat_size = sizeof(CXLMemPatrolScrubWriteAttrbs),
> +                /* Bit[0] : 1, feature attributes changeable */
> +                .attrb_flags = 0x1,
> +                .get_feat_version = CXL_MEMDEV_PS_GET_FEATURE_VERSION,
> +                .set_feat_version = CXL_MEMDEV_PS_SET_FEATURE_VERSION,
> +                .set_feat_effects = 0,
> +            };
> +            feat_entries++;
> +            /* Set default value for device patrol scrub read attributes */
> +            cxl_memdev_ps_feat_read_attrbs.scrub_cycle_cap =
> +                                CXL_MEMDEV_PS_SCRUB_CYCLE_CHANGE_CAP_DEFAULT |
> +                                CXL_MEMDEV_PS_SCRUB_REALTIME_REPORT_CAP_DEFAULT;
> +            cxl_memdev_ps_feat_read_attrbs.scrub_cycle =
> +                                CXL_MEMDEV_PS_CUR_SCRUB_CYCLE_DEFAULT |
> +                                (CXL_MEMDEV_PS_MIN_SCRUB_CYCLE_DEFAULT << 8);
> +            cxl_memdev_ps_feat_read_attrbs.scrub_flags =
> +                                CXL_MEMDEV_PS_ENABLE_DEFAULT;
> +            break;
>          default:
>              break;
>          }
> @@ -924,6 +981,21 @@ static CXLRetCode cmd_features_get_feature(const struct cxl_cmd *cmd,
>          return CXL_MBOX_INVALID_INPUT;
>      }
>  
> +    if (qemu_uuid_is_equal(&get_feature->uuid, &patrol_scrub_uuid)) {
> +        if (get_feature->offset >= sizeof(cxl_memdev_ps_feat_read_attrbs)) {
> +            return CXL_MBOX_INVALID_INPUT;
> +        }
> +        bytes_to_copy = sizeof(cxl_memdev_ps_feat_read_attrbs) -
> +                                             get_feature->offset;
> +        bytes_to_copy = (bytes_to_copy > get_feature->count) ?
> +                               get_feature->count : bytes_to_copy;
> +        memcpy(payload_out,
> +               &cxl_memdev_ps_feat_read_attrbs + get_feature->offset,
> +               bytes_to_copy);
> +    } else {
> +        return CXL_MBOX_UNSUPPORTED;
> +    }
> +
>      *len_out = bytes_to_copy;
>  
>      return CXL_MBOX_SUCCESS;
> @@ -937,6 +1009,29 @@ static CXLRetCode cmd_features_set_feature(const struct cxl_cmd *cmd,
>                                             size_t *len_out,
>                                             CXLCCI *cci)
>  {
> +    CXLMemPatrolScrubWriteAttrbs *ps_write_attrbs;
> +    CXLMemPatrolScrubSetFeature *ps_set_feature;
> +    CXLSetFeatureInHeader *hdr = (void *)payload_in;
> +
> +    if (qemu_uuid_is_equal(&hdr->uuid, &patrol_scrub_uuid)) {
> +        if (hdr->version != CXL_MEMDEV_PS_SET_FEATURE_VERSION ||
> +            (hdr->flags & CXL_SET_FEATURE_FLAG_DATA_TRANSFER_MASK) !=
> +                               CXL_SET_FEATURE_FLAG_FULL_DATA_TRANSFER) {
> +            return CXL_MBOX_UNSUPPORTED;
> +        }
> +
> +        ps_set_feature = (void *)payload_in;
> +        ps_write_attrbs = &ps_set_feature->feat_data;
> +        cxl_memdev_ps_feat_read_attrbs.scrub_cycle &= ~0xFF;
> +        cxl_memdev_ps_feat_read_attrbs.scrub_cycle |=
> +                          ps_write_attrbs->scrub_cycle_hr & 0xFF;
> +        cxl_memdev_ps_feat_read_attrbs.scrub_flags &= ~0x1;
> +        cxl_memdev_ps_feat_read_attrbs.scrub_flags |=
> +                          ps_write_attrbs->scrub_flags & 0x1;
> +    } else {
> +        return CXL_MBOX_UNSUPPORTED;
> +    }
> +
>      return CXL_MBOX_SUCCESS;
>  }
>  
> -- 
> 2.34.1
> 

  reply	other threads:[~2024-02-15 18:47 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-24 13:53 [PATCH v2 0/3] hw/cxl/cxl-mailbox-utils: Add feature commands, device patrol scrub control and DDR5 ECS control features shiju.jose
2023-11-24 13:53 ` shiju.jose--- via
2023-11-24 13:53 ` [PATCH v2 1/3] hw/cxl/cxl-mailbox-utils: Add support for feature commands (8.2.9.6) shiju.jose
2023-11-24 13:53   ` shiju.jose--- via
2024-02-15 18:05   ` fan
2024-02-16  9:20     ` Shiju Jose
2024-02-16  9:20       ` Shiju Jose via
2023-11-24 13:53 ` [PATCH v2 2/3] hw/cxl/cxl-mailbox-utils: Add device patrol scrub control feature shiju.jose
2023-11-24 13:53   ` shiju.jose--- via
2024-02-15 18:47   ` fan [this message]
2024-02-16 10:16     ` Shiju Jose
2024-02-16 10:16       ` Shiju Jose via
2024-02-16 17:35       ` fan
2024-02-15 20:55   ` Davidlohr Bueso
2024-02-16 12:11     ` Shiju Jose
2024-02-16 12:11       ` Shiju Jose via
2023-11-24 13:53 ` [PATCH v2 3/3] hw/cxl/cxl-mailbox-utils: Add device DDR5 ECS " shiju.jose
2023-11-24 13:53   ` shiju.jose--- via
2024-02-15 19:11   ` fan
2024-02-15 18:08 ` [PATCH v2 0/3] hw/cxl/cxl-mailbox-utils: Add feature commands, device patrol scrub control and DDR5 ECS control features fan
2024-02-16  9:22   ` Shiju Jose
2024-02-16  9:22     ` Shiju Jose via

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=Zc5cNLH9FS24qXTe@debian \
    --to=nifan.cxl@gmail.com \
    --cc=fan.ni@samsung.com \
    --cc=jonathan.cameron@huawei.com \
    --cc=linux-cxl@vger.kernel.org \
    --cc=linuxarm@huawei.com \
    --cc=prime.zeng@hisilicon.com \
    --cc=qemu-devel@nongnu.org \
    --cc=shiju.jose@huawei.com \
    --cc=tanxiaofei@huawei.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.