From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by dpdk.org (Postfix) with ESMTP id 1E96F1B67E for ; Wed, 9 May 2018 16:18:01 +0200 (CEST) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga001.fm.intel.com ([10.253.24.23]) by orsmga102.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 09 May 2018 07:18:00 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.49,381,1520924400"; d="scan'208";a="53725922" Received: from kchalupx-mobl.ger.corp.intel.com ([10.103.104.151]) by fmsmga001.fm.intel.com with ESMTP; 09 May 2018 07:17:59 -0700 From: Kamil Chalupnik To: dev@dpdk.org Cc: amr.mokhtar@intel.com, KamilX Chalupnik Date: Wed, 9 May 2018 16:17:56 +0200 Message-Id: <20180509141756.23604-1-kamilx.chalupnik@intel.com> X-Mailer: git-send-email 2.9.0.windows.1 In-Reply-To: <20180426133008.12388-4-kamilx.chalupnik@intel.com> References: <20180426133008.12388-4-kamilx.chalupnik@intel.com> Subject: [dpdk-dev] [PATCH v2 02/14] baseband/turbo_sw: memory copying optimized or removed X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 09 May 2018 14:18:02 -0000 From: KamilX Chalupnik Optimization of Turbo Software driver by limiting or changing usage of memory copying function. Signed-off-by: Kamil Chalupnik --- drivers/baseband/turbo_sw/bbdev_turbo_software.c | 100 +++++++++++++++-------- 1 file changed, 64 insertions(+), 36 deletions(-) diff --git a/drivers/baseband/turbo_sw/bbdev_turbo_software.c b/drivers/baseband/turbo_sw/bbdev_turbo_software.c index 26b8560..35d2827 100644 --- a/drivers/baseband/turbo_sw/bbdev_turbo_software.c +++ b/drivers/baseband/turbo_sw/bbdev_turbo_software.c @@ -452,7 +452,7 @@ is_dec_input_valid(int32_t k_idx, int16_t kw, int16_t in_length) static inline void process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, - uint8_t cb_idx, uint8_t c, uint16_t k, uint16_t ncb, + uint8_t r, uint8_t c, uint16_t k, uint16_t ncb, uint32_t e, struct rte_mbuf *m_in, struct rte_mbuf *m_out, uint16_t in_offset, uint16_t out_offset, uint16_t total_left) { @@ -460,6 +460,7 @@ process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, int16_t k_idx; uint16_t m; uint8_t *in, *out0, *out1, *out2, *tmp_out, *rm_out; + uint64_t first_3_bytes = 0; struct rte_bbdev_op_turbo_enc *enc = &op->turbo_enc; struct bblib_crc_request crc_req; struct bblib_crc_response crc_resp; @@ -479,16 +480,25 @@ process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, op->status |= 1 << RTE_BBDEV_DATA_ERROR; return; } - /* copy the input to the temporary buffer to be able to extend - * it by 3 CRC bytes - */ - rte_memcpy(q->enc_in, in, (k - 24) >> 3); crc_req.data = in; crc_req.len = (k - 24) >> 3; - crc_resp.data = q->enc_in; - bblib_lte_crc24a_gen(&crc_req, &crc_resp); + /* Check if there is a room for CRC bits. If not use + * the temporary buffer. + */ + if (rte_pktmbuf_append(m_in, 3) == NULL) { + rte_memcpy(q->enc_in, in, (k - 24) >> 3); + in = q->enc_in; + } else { + /* Store 3 first bytes of next CB as they will be + * overwritten by CRC bytes. If it is the last CB then + * there is no point to store 3 next bytes and this + * if..else branch will be omitted. + */ + first_3_bytes = *((uint64_t *)&in[(k - 32) >> 3]); + } - in = q->enc_in; + crc_resp.data = in; + bblib_lte_crc24a_gen(&crc_req, &crc_resp); } else if (enc->op_flags & RTE_BBDEV_TURBO_CRC_24B_ATTACH) { /* CRC24B */ ret = is_enc_input_valid(k - 24, k_idx, total_left); @@ -496,16 +506,25 @@ process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, op->status |= 1 << RTE_BBDEV_DATA_ERROR; return; } - /* copy the input to the temporary buffer to be able to extend - * it by 3 CRC bytes - */ - rte_memcpy(q->enc_in, in, (k - 24) >> 3); crc_req.data = in; crc_req.len = (k - 24) >> 3; - crc_resp.data = q->enc_in; - bblib_lte_crc24b_gen(&crc_req, &crc_resp); + /* Check if there is a room for CRC bits. If this is the last + * CB in TB. If not use temporary buffer. + */ + if ((c - r == 1) && (rte_pktmbuf_append(m_in, 3) == NULL)) { + rte_memcpy(q->enc_in, in, (k - 24) >> 3); + in = q->enc_in; + } else if (c - r > 1) { + /* Store 3 first bytes of next CB as they will be + * overwritten by CRC bytes. If it is the last CB then + * there is no point to store 3 next bytes and this + * if..else branch will be omitted. + */ + first_3_bytes = *((uint64_t *)&in[(k - 32) >> 3]); + } - in = q->enc_in; + crc_resp.data = in; + bblib_lte_crc24b_gen(&crc_req, &crc_resp); } else { ret = is_enc_input_valid(k, k_idx, total_left); if (ret != 0) { @@ -519,10 +538,32 @@ process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, /* Each bit layer output from turbo encoder is (k+4) bits long, i.e. * input length + 4 tail bits. That's (k/8) + 1 bytes after rounding up. * So dst_data's length should be 3*(k/8) + 3 bytes. + * In Rate-matching bypass case outputs pointers passed to encoder + * (out0, out1 and out2) can directly point to addresses of output from + * turbo_enc entity. */ - out0 = q->enc_out; - out1 = RTE_PTR_ADD(out0, (k >> 3) + 1); - out2 = RTE_PTR_ADD(out1, (k >> 3) + 1); + if (enc->op_flags & RTE_BBDEV_TURBO_RATE_MATCH) { + out0 = q->enc_out; + out1 = RTE_PTR_ADD(out0, (k >> 3) + 1); + out2 = RTE_PTR_ADD(out1, (k >> 3) + 1); + } else { + out0 = (uint8_t *)rte_pktmbuf_append(m_out, (k >> 3) * 3 + 2); + if (out0 == NULL) { + op->status |= 1 << RTE_BBDEV_DATA_ERROR; + rte_bbdev_log(ERR, + "Too little space in output mbuf"); + return; + } + enc->output.length += (k >> 3) * 3 + 2; + /* rte_bbdev_op_data.offset can be different than the + * offset of the appended bytes + */ + out0 = rte_pktmbuf_mtod_offset(m_out, uint8_t *, out_offset); + out1 = rte_pktmbuf_mtod_offset(m_out, uint8_t *, + out_offset + (k >> 3) + 1); + out2 = rte_pktmbuf_mtod_offset(m_out, uint8_t *, + out_offset + 2 * ((k >> 3) + 1)); + } turbo_req.case_id = k_idx; turbo_req.input_win = in; @@ -536,6 +577,10 @@ process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, return; } + /* Restore 3 first bytes of next CB if they were overwritten by CRC*/ + if (first_3_bytes != 0) + *((uint64_t *)&in[(k - 32) >> 3]) = first_3_bytes; + /* Rate-matching */ if (enc->op_flags & RTE_BBDEV_TURBO_RATE_MATCH) { /* get output data starting address */ @@ -552,7 +597,7 @@ process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, rm_out = rte_pktmbuf_mtod_offset(m_out, uint8_t *, out_offset); /* index of current code block */ - rm_req.r = cb_idx; + rm_req.r = r; /* total number of code block */ rm_req.C = c; /* For DL - 1, UL - 0 */ @@ -613,23 +658,6 @@ process_enc_cb(struct turbo_sw_queue *q, struct rte_bbdev_enc_op *op, tmp_out++; } *tmp_out = 0; - - /* copy shifted output to turbo_enc entity */ - out0 = (uint8_t *)rte_pktmbuf_append(m_out, - (k >> 3) * 3 + 2); - if (out0 == NULL) { - op->status |= 1 << RTE_BBDEV_DATA_ERROR; - rte_bbdev_log(ERR, - "Too little space in output mbuf"); - return; - } - enc->output.length += (k >> 3) * 3 + 2; - /* rte_bbdev_op_data.offset can be different than the - * offset of the appended bytes - */ - out0 = rte_pktmbuf_mtod_offset(m_out, uint8_t *, - out_offset); - rte_memcpy(out0, q->enc_out, (k >> 3) * 3 + 2); } } -- 2.5.5