From: Leon Romanovsky <leon@kernel.org>
To: Selvin Xavier <selvin.xavier@broadcom.com>
Cc: jgg@ziepe.ca, dledford@redhat.com, linux-rdma@vger.kernel.org,
Naresh Kumar PBS <nareshkumar.pbs@broadcom.com>
Subject: Re: [PATCH rdma-rc] RDMA/bnxt_re: Fix stats counters
Date: Sun, 11 Jul 2021 15:00:32 +0300 [thread overview]
Message-ID: <YOrdYGyJu/n6yWJ2@unreal> (raw)
In-Reply-To: <1625992490-5127-1-git-send-email-selvin.xavier@broadcom.com>
On Sun, Jul 11, 2021 at 01:34:50AM -0700, Selvin Xavier wrote:
> From: Naresh Kumar PBS <nareshkumar.pbs@broadcom.com>
>
> Stats counters are not incrementing in some adapter versions
> with newer FW. This is due to the stats context length mismatch
> between FW and driver. Since L2 driver updates the length correctly,
> use the stats length from L2 driver while allocating the DMA'able
> memory and creating the stats context.
>
> Fixes:9d6b648c311(bnxt_en: Update firmware interface spec to 1.10.1.65)
This is wrong fixes line.
I recommend to add the line below to your gitaliases file
fixes = "!git --no-pager log --abbrev=12 -1 --format='Fixes: %h (\"%s\")'"
and it will give you nice git fixes command with ready to copy string:
➜ kernel git:(rdma-next) git fixes 9d6b648c311
Fixes: 9d6b648c3112 ("bnxt_en: Update firmware interface spec to 1.10.1.65.")
> Signed-off-by: Naresh Kumar PBS <nareshkumar.pbs@broadcom.com>
> Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
> ---
> drivers/infiniband/hw/bnxt_re/main.c | 4 +++-
> drivers/infiniband/hw/bnxt_re/qplib_res.c | 10 ++++------
> drivers/infiniband/hw/bnxt_re/qplib_res.h | 1 +
> 3 files changed, 8 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/infiniband/hw/bnxt_re/main.c b/drivers/infiniband/hw/bnxt_re/main.c
> index 04621ba..1fadca8 100644
> --- a/drivers/infiniband/hw/bnxt_re/main.c
> +++ b/drivers/infiniband/hw/bnxt_re/main.c
> @@ -119,6 +119,7 @@ static int bnxt_re_setup_chip_ctx(struct bnxt_re_dev *rdev, u8 wqe_mode)
> if (!chip_ctx)
> return -ENOMEM;
> chip_ctx->chip_num = bp->chip_num;
> + chip_ctx->hw_stats_size = bp->hw_ring_stats_size;
>
> rdev->chip_ctx = chip_ctx;
> /* rest members to follow eventually */
> @@ -507,6 +508,7 @@ static int bnxt_re_net_stats_ctx_alloc(struct bnxt_re_dev *rdev,
> dma_addr_t dma_map,
> u32 *fw_stats_ctx_id)
> {
> + struct bnxt_qplib_chip_ctx *chip_ctx = rdev->chip_ctx;
> struct hwrm_stat_ctx_alloc_output resp = {0};
> struct hwrm_stat_ctx_alloc_input req = {0};
> struct bnxt_en_dev *en_dev = rdev->en_dev;
> @@ -523,7 +525,7 @@ static int bnxt_re_net_stats_ctx_alloc(struct bnxt_re_dev *rdev,
> bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_STAT_CTX_ALLOC, -1, -1);
> req.update_period_ms = cpu_to_le32(1000);
> req.stats_dma_addr = cpu_to_le64(dma_map);
> - req.stats_dma_length = cpu_to_le16(sizeof(struct ctx_hw_stats_ext));
> + req.stats_dma_length = cpu_to_le16(chip_ctx->hw_stats_size);
> req.stat_ctx_flags = STAT_CTX_ALLOC_REQ_STAT_CTX_FLAGS_ROCE;
> bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp,
> sizeof(resp), DFLT_HWRM_CMD_TIMEOUT);
> diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.c b/drivers/infiniband/hw/bnxt_re/qplib_res.c
> index fa78783..72be4fb 100644
> --- a/drivers/infiniband/hw/bnxt_re/qplib_res.c
> +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.c
> @@ -56,6 +56,7 @@
> static void bnxt_qplib_free_stats_ctx(struct pci_dev *pdev,
> struct bnxt_qplib_stats *stats);
> static int bnxt_qplib_alloc_stats_ctx(struct pci_dev *pdev,
> + struct bnxt_qplib_chip_ctx *cctx,
> struct bnxt_qplib_stats *stats);
>
> /* PBL */
> @@ -559,7 +560,7 @@ int bnxt_qplib_alloc_ctx(struct bnxt_qplib_res *res,
> goto fail;
> stats_alloc:
> /* Stats */
> - rc = bnxt_qplib_alloc_stats_ctx(res->pdev, &ctx->stats);
> + rc = bnxt_qplib_alloc_stats_ctx(res->pdev, res->cctx, &ctx->stats);
> if (rc)
> goto fail;
>
> @@ -888,15 +889,12 @@ static void bnxt_qplib_free_stats_ctx(struct pci_dev *pdev,
> }
>
> static int bnxt_qplib_alloc_stats_ctx(struct pci_dev *pdev,
> + struct bnxt_qplib_chip_ctx *cctx,
> struct bnxt_qplib_stats *stats)
> {
> memset(stats, 0, sizeof(*stats));
> stats->fw_id = -1;
> - /* 128 byte aligned context memory is required only for 57500.
> - * However making this unconditional, it does not harm previous
> - * generation.
> - */
> - stats->size = ALIGN(sizeof(struct ctx_hw_stats), 128);
> + stats->size = cctx->hw_stats_size;
> stats->dma = dma_alloc_coherent(&pdev->dev, stats->size,
> &stats->dma_map, GFP_KERNEL);
> if (!stats->dma) {
> diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.h b/drivers/infiniband/hw/bnxt_re/qplib_res.h
> index 7a1ab38..58bad6f 100644
> --- a/drivers/infiniband/hw/bnxt_re/qplib_res.h
> +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.h
> @@ -60,6 +60,7 @@ struct bnxt_qplib_chip_ctx {
> u16 chip_num;
> u8 chip_rev;
> u8 chip_metal;
> + u16 hw_stats_size;
> struct bnxt_qplib_drv_modes modes;
> };
>
> --
> 2.5.5
>
next prev parent reply other threads:[~2021-07-11 12:00 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-07-11 8:34 [PATCH rdma-rc] RDMA/bnxt_re: Fix stats counters Selvin Xavier
2021-07-11 12:00 ` Leon Romanovsky [this message]
2021-07-11 13:29 ` Selvin Xavier
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=YOrdYGyJu/n6yWJ2@unreal \
--to=leon@kernel.org \
--cc=dledford@redhat.com \
--cc=jgg@ziepe.ca \
--cc=linux-rdma@vger.kernel.org \
--cc=nareshkumar.pbs@broadcom.com \
--cc=selvin.xavier@broadcom.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.