From patchwork Thu Nov 9 16:35:03 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Benjamin Gaignard X-Patchwork-Id: 13451547 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 8B508C4167D for ; Thu, 9 Nov 2023 16:40:57 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:MIME-Version:References:In-Reply-To: Message-Id:Date:Subject:Cc:To:From:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=BBZHdHSAnh9e+dFEi1jXr0XjCKY6ct4Y+UBL4tWhdXA=; b=T/23mwr/K9lREM uhgfLWKWg//IzEtAf0r7CG6n6FncU2Y6nrAwEb3QuQ3JBUjpDFliIma7qqba3ugJzcWp61Djcjara NhU2f+GCLyx7lpxgGVPcc3h1j7IJDCJtAsw3GvCW9LX+P9oL3+s08kf1bTZ5AO1yv2vQV3hx1iQwQ wzTydIh9e+sS9cHoAYdZoXMsv5fTrgSm34KTaAX5yiVD/1cHH+G7EnINM+ODovGFvboty8vdO81NU oxp+fyAGOfOcX48+4XVMqMAQBAF0LTXLzrAR52IeD0Y3eYcG37VTUUnsnIGGSy6nY280TDJklIVFj fe//9AM9udPNdCa9eR9Q==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.96 #2 (Red Hat Linux)) id 1r1850-006o9d-2p; Thu, 09 Nov 2023 16:40:30 +0000 Received: from madras.collabora.co.uk ([46.235.227.172]) by bombadil.infradead.org with esmtps (Exim 4.96 #2 (Red Hat Linux)) id 1r183C-006m15-0t; Thu, 09 Nov 2023 16:38:47 +0000 Received: from benjamin-XPS-13-9310.. (cola.collaboradmins.com [195.201.22.229]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) (Authenticated sender: benjamin.gaignard) by madras.collabora.co.uk (Postfix) with ESMTPSA id D6D5A66076C5; Thu, 9 Nov 2023 16:38:35 +0000 (GMT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1699547916; bh=YtbmcbhNvbDBINI9lxYBaiInifnOxrUmV+MuBuuk8ew=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=gRXqlgIFvdUsmX/IG7E0BWpswcKKMjSmmb3+kedqXV/hSXQq8ubnch5Y7Bpa9tej/ ErkeWbMF3aR6hrcWvVo7MF8fhEGtcBvwXCnzpw6Q9ZejT+VJiTXdxKWJn36zqHfhSg Q4C3lhz96Ta1ayyVDk7F7nhEI6LF7nyXjEGJgar8AlfsR4sNFRqDOD8Uj6iD0uwyHy c2klKRfSurbTnDrqqFioALpEGSn/axm2ZnraDj87w38/PYdlVEjNABl9rZp90gxERg nlms4jycOxN7V52dznz0Eac1aYk9n6Eu5CouewEg9u8lmaY+urYCwbl/3Si6m5rQ1Q h7I6ar1TLEx9g== From: Benjamin Gaignard To: mchehab@kernel.org, tfiga@chromium.org, m.szyprowski@samsung.com, ming.qian@nxp.com, ezequiel@vanguardiasur.com.ar, p.zabel@pengutronix.de, gregkh@linuxfoundation.org, hverkuil-cisco@xs4all.nl, nicolas.dufresne@collabora.com Cc: linux-media@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-mediatek@lists.infradead.org, linux-arm-msm@vger.kernel.org, linux-rockchip@lists.infradead.org, linux-staging@lists.linux.dev, kernel@collabora.com, Benjamin Gaignard , Andrzej Pietrasiewicz Subject: [PATCH v15 47/56] media: verisilicon: Refactor postprocessor to store more buffers Date: Thu, 9 Nov 2023 17:35:03 +0100 Message-Id: <20231109163512.179524-48-benjamin.gaignard@collabora.com> X-Mailer: git-send-email 2.39.2 In-Reply-To: <20231109163512.179524-14-benjamin.gaignard@collabora.com> References: <20231109163512.179524-14-benjamin.gaignard@collabora.com> MIME-Version: 1.0 X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20231109_083838_587082_F39EC649 X-CRM114-Status: GOOD ( 22.36 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org Since vb2 queue can store more than VB2_MAX_FRAME buffers, the postprocessor buffer storage must be capable to store more buffers too. Change static dec_q array to allocated array to be capable to store up to queue 'max_num_buffers'. Keep allocating queue 'num_buffers' at queue setup time but also allows to allocate postprocessors buffers on the fly. Signed-off-by: Benjamin Gaignard Reviewed-by: Andrzej Pietrasiewicz CC: Ezequiel Garcia CC: Philipp Zabel --- drivers/media/platform/verisilicon/hantro.h | 7 +- .../media/platform/verisilicon/hantro_drv.c | 4 +- .../media/platform/verisilicon/hantro_hw.h | 4 +- .../platform/verisilicon/hantro_postproc.c | 93 +++++++++++++++---- .../media/platform/verisilicon/hantro_v4l2.c | 2 +- 5 files changed, 85 insertions(+), 25 deletions(-) diff --git a/drivers/media/platform/verisilicon/hantro.h b/drivers/media/platform/verisilicon/hantro.h index 77aee9489516..0948b04a9f8d 100644 --- a/drivers/media/platform/verisilicon/hantro.h +++ b/drivers/media/platform/verisilicon/hantro.h @@ -469,11 +469,14 @@ hantro_get_dst_buf(struct hantro_ctx *ctx) bool hantro_needs_postproc(const struct hantro_ctx *ctx, const struct hantro_fmt *fmt); +dma_addr_t +hantro_postproc_get_dec_buf_addr(struct hantro_ctx *ctx, int index); + static inline dma_addr_t hantro_get_dec_buf_addr(struct hantro_ctx *ctx, struct vb2_buffer *vb) { if (hantro_needs_postproc(ctx, ctx->vpu_dst_fmt)) - return ctx->postproc.dec_q[vb->index].dma; + return hantro_postproc_get_dec_buf_addr(ctx, vb->index); return vb2_dma_contig_plane_dma_addr(vb, 0); } @@ -485,8 +488,8 @@ vb2_to_hantro_decoded_buf(struct vb2_buffer *buf) void hantro_postproc_disable(struct hantro_ctx *ctx); void hantro_postproc_enable(struct hantro_ctx *ctx); +int hantro_postproc_init(struct hantro_ctx *ctx); void hantro_postproc_free(struct hantro_ctx *ctx); -int hantro_postproc_alloc(struct hantro_ctx *ctx); int hanto_postproc_enum_framesizes(struct hantro_ctx *ctx, struct v4l2_frmsizeenum *fsize); diff --git a/drivers/media/platform/verisilicon/hantro_drv.c b/drivers/media/platform/verisilicon/hantro_drv.c index a9fa05ac56a9..7f5b82eb6649 100644 --- a/drivers/media/platform/verisilicon/hantro_drv.c +++ b/drivers/media/platform/verisilicon/hantro_drv.c @@ -235,8 +235,10 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq) * The Kernel needs access to the JPEG destination buffer for the * JPEG encoder to fill in the JPEG headers. */ - if (!ctx->is_encoder) + if (!ctx->is_encoder) { dst_vq->dma_attrs |= DMA_ATTR_NO_KERNEL_MAPPING; + dst_vq->max_num_buffers = MAX_POSTPROC_BUFFERS; + } dst_vq->type = V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE; dst_vq->io_modes = VB2_MMAP | VB2_DMABUF; diff --git a/drivers/media/platform/verisilicon/hantro_hw.h b/drivers/media/platform/verisilicon/hantro_hw.h index 7f33f7b07ce4..292a76ef643e 100644 --- a/drivers/media/platform/verisilicon/hantro_hw.h +++ b/drivers/media/platform/verisilicon/hantro_hw.h @@ -40,6 +40,8 @@ #define AV1_MAX_FRAME_BUF_COUNT (V4L2_AV1_TOTAL_REFS_PER_FRAME + 1) +#define MAX_POSTPROC_BUFFERS 64 + struct hantro_dev; struct hantro_ctx; struct hantro_buf; @@ -336,7 +338,7 @@ struct hantro_av1_dec_hw_ctx { * @dec_q: References buffers, in decoder format. */ struct hantro_postproc_ctx { - struct hantro_aux_buf dec_q[VB2_MAX_FRAME]; + struct hantro_aux_buf dec_q[MAX_POSTPROC_BUFFERS]; }; /** diff --git a/drivers/media/platform/verisilicon/hantro_postproc.c b/drivers/media/platform/verisilicon/hantro_postproc.c index 8f8f17e671ce..41e93176300b 100644 --- a/drivers/media/platform/verisilicon/hantro_postproc.c +++ b/drivers/media/platform/verisilicon/hantro_postproc.c @@ -177,9 +177,11 @@ static int hantro_postproc_g2_enum_framesizes(struct hantro_ctx *ctx, void hantro_postproc_free(struct hantro_ctx *ctx) { struct hantro_dev *vpu = ctx->dev; + struct v4l2_m2m_ctx *m2m_ctx = ctx->fh.m2m_ctx; + struct vb2_queue *queue = &m2m_ctx->cap_q_ctx.q; unsigned int i; - for (i = 0; i < VB2_MAX_FRAME; ++i) { + for (i = 0; i < queue->max_num_buffers; ++i) { struct hantro_aux_buf *priv = &ctx->postproc.dec_q[i]; if (priv->cpu) { @@ -190,20 +192,17 @@ void hantro_postproc_free(struct hantro_ctx *ctx) } } -int hantro_postproc_alloc(struct hantro_ctx *ctx) +static unsigned int hantro_postproc_buffer_size(struct hantro_ctx *ctx) { - struct hantro_dev *vpu = ctx->dev; - struct v4l2_m2m_ctx *m2m_ctx = ctx->fh.m2m_ctx; - struct vb2_queue *cap_queue = &m2m_ctx->cap_q_ctx.q; - unsigned int num_buffers = vb2_get_num_buffers(cap_queue); struct v4l2_pix_format_mplane pix_mp; const struct hantro_fmt *fmt; - unsigned int i, buf_size; + unsigned int buf_size; /* this should always pick native format */ fmt = hantro_get_default_fmt(ctx, false, ctx->bit_depth, HANTRO_AUTO_POSTPROC); if (!fmt) - return -EINVAL; + return 0; + v4l2_fill_pixfmt_mp(&pix_mp, fmt->fourcc, ctx->src_fmt.width, ctx->src_fmt.height); @@ -221,23 +220,77 @@ int hantro_postproc_alloc(struct hantro_ctx *ctx) buf_size += hantro_av1_mv_size(pix_mp.width, pix_mp.height); - for (i = 0; i < num_buffers; ++i) { - struct hantro_aux_buf *priv = &ctx->postproc.dec_q[i]; + return buf_size; +} + +static int hantro_postproc_alloc(struct hantro_ctx *ctx, int index) +{ + struct hantro_dev *vpu = ctx->dev; + struct hantro_aux_buf *priv = &ctx->postproc.dec_q[index]; + unsigned int buf_size = hantro_postproc_buffer_size(ctx); + + if (!buf_size) + return -EINVAL; + + /* + * The buffers on this queue are meant as intermediate + * buffers for the decoder, so no mapping is needed. + */ + priv->attrs = DMA_ATTR_NO_KERNEL_MAPPING; + priv->cpu = dma_alloc_attrs(vpu->dev, buf_size, &priv->dma, + GFP_KERNEL, priv->attrs); + if (!priv->cpu) + return -ENOMEM; + priv->size = buf_size; + + return 0; +} - /* - * The buffers on this queue are meant as intermediate - * buffers for the decoder, so no mapping is needed. - */ - priv->attrs = DMA_ATTR_NO_KERNEL_MAPPING; - priv->cpu = dma_alloc_attrs(vpu->dev, buf_size, &priv->dma, - GFP_KERNEL, priv->attrs); - if (!priv->cpu) - return -ENOMEM; - priv->size = buf_size; +int hantro_postproc_init(struct hantro_ctx *ctx) +{ + struct v4l2_m2m_ctx *m2m_ctx = ctx->fh.m2m_ctx; + struct vb2_queue *cap_queue = &m2m_ctx->cap_q_ctx.q; + unsigned int num_buffers = vb2_get_num_buffers(cap_queue); + unsigned int i; + int ret; + + for (i = 0; i < num_buffers; i++) { + ret = hantro_postproc_alloc(ctx, i); + if (ret) + return ret; } + return 0; } +dma_addr_t +hantro_postproc_get_dec_buf_addr(struct hantro_ctx *ctx, int index) +{ + struct hantro_aux_buf *priv = &ctx->postproc.dec_q[index]; + unsigned int buf_size = hantro_postproc_buffer_size(ctx); + struct hantro_dev *vpu = ctx->dev; + int ret; + + if (priv->size < buf_size && priv->cpu) { + /* buffer is too small, release it */ + dma_free_attrs(vpu->dev, priv->size, priv->cpu, + priv->dma, priv->attrs); + priv->cpu = NULL; + } + + if (!priv->cpu) { + /* buffer not already allocated, try getting a new one */ + ret = hantro_postproc_alloc(ctx, index); + if (ret) + return 0; + } + + if (!priv->cpu) + return 0; + + return priv->dma; +} + static void hantro_postproc_g1_disable(struct hantro_ctx *ctx) { struct hantro_dev *vpu = ctx->dev; diff --git a/drivers/media/platform/verisilicon/hantro_v4l2.c b/drivers/media/platform/verisilicon/hantro_v4l2.c index b3ae037a50f6..f0d8b165abcd 100644 --- a/drivers/media/platform/verisilicon/hantro_v4l2.c +++ b/drivers/media/platform/verisilicon/hantro_v4l2.c @@ -933,7 +933,7 @@ static int hantro_start_streaming(struct vb2_queue *q, unsigned int count) } if (hantro_needs_postproc(ctx, ctx->vpu_dst_fmt)) { - ret = hantro_postproc_alloc(ctx); + ret = hantro_postproc_init(ctx); if (ret) goto err_codec_exit; }