From mboxrd@z Thu Jan 1 00:00:00 1970 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 mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9E25AD46BE5 for ; Wed, 28 Jan 2026 17:52:37 +0000 (UTC) Received: from mails.dpdk.org (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id F2DA44026F; Wed, 28 Jan 2026 18:52:36 +0100 (CET) Received: from mail-wr1-f46.google.com (mail-wr1-f46.google.com [209.85.221.46]) by mails.dpdk.org (Postfix) with ESMTP id 707894003C for ; Wed, 28 Jan 2026 18:52:35 +0100 (CET) Received: by mail-wr1-f46.google.com with SMTP id ffacd0b85a97d-42fb4eeb482so122883f8f.0 for ; Wed, 28 Jan 2026 09:52:35 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=networkplumber-org.20230601.gappssmtp.com; s=20230601; t=1769622755; x=1770227555; darn=dpdk.org; h=mime-version:content-transfer-encoding:references:in-reply-to :subject:to:from:date:message-id:from:to:cc:subject:date:message-id :reply-to; bh=0bkj4HwH27cJyrtPsgVwKS1LwlMu1Bq0NPY63D9/cU4=; b=C4OE+c25SUnbWlCDZwal0e24bzNA82JOnLC7W+ttpxkFnmrk4qUTQsG6EZmr3sPDMt KsBv0jkdsTGyk5lUBEwyX+kxslrRY8dWQG5yQ0qt21Rd9AANaYCJsSUPLkyjq3D9yIvs wWcj7GSI75rWJmUsWHEdqJ4Br3BlUxKwgQ/xqB7qsL+++pAEIk+XGTdbMz9mOQde6WYh wjeGmJ9q1d9F/w5wwPyttSahoeVEEtXkJ1pzyUlnFmdNm+3JJLZVkM9ARGMqaUGJ1V4M M6QF3z4th6OeuJz7yrngflA0tAumZvPRnXpy7QsFy6qdvQOH1cwqpc4PMGpLWs6SEMsh 72ZA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1769622755; x=1770227555; h=mime-version:content-transfer-encoding:references:in-reply-to :subject:to:from:date:message-id:x-gm-gg:x-gm-message-state:from:to :cc:subject:date:message-id:reply-to; bh=0bkj4HwH27cJyrtPsgVwKS1LwlMu1Bq0NPY63D9/cU4=; b=Tso9JSliXOavO3w8x3nwz0+TxfQw2+rxwFl44TKvHeogXCn9ZFmnH+vQt4xr0Jlvn8 5K9zqncnu8Vqaz/PZEKa9CysjGrFDGSOt5RMBf80XtW530hy553+gjVD3MtKp7YCRu+v I9ERTvobUvBoUxs4ZmloR7nIgnLmE1nJvxJED0Xq4wNVvvptQ/faRkmalol1UyjF3WIs qX3FZYIwP/ZD0dYb+pqIFwP+KmEcLdcxUcKwaPVKYDQ9rPJcRsNDecJ4wfizFO2JpQjX YoMINZWln1erKSw49Ff7/1d86vTMgqxTDzMTSgqBK8b0oozfd15LG2jLpPmlSPZbdHyS +4JA== X-Gm-Message-State: AOJu0YycQ4WLIAcVNx8HaFgt0sy71wBfRUi/HQiIYKEGlHbfAXyXUAoD 7oW9cKZM39fV0i4YP64Iu02MDh3b27JPEjzpIvHGoAjv4TcY/AUpe99mAczjw3IUO1Lu+mLi8nJ +fm9g X-Gm-Gg: AZuq6aI0J1lynEXhdXBUVU6JQukYE2TFvx+qX+bUNwOUG89/RL2aHv1X6vfxue/3U6u y2aJasLVCw8zCb5XCjDV8Cr0mFo/cn4tX6FpBPR68gsc4i60FpEVXbDzETwRSw7FeyHneYecz7w 67f0swmxVyplIH1/XuEzbnSywBGfQRGumKK10OLHiff9qhCzmcgNrOzflR2WTedoCHvXu+vJgCF KJwUXNdMflxIrKp0Y2GXV5Ltmooj12Z/pRkNkHPlsmc+Mc8KU0ThE652bPMEU/Jw0pmJaOYC6XU ESZnVc8rlGl/aOhhrcsuGz6VV2fH3DfaJotSkV1mUPMyjOGN4JMHUai4stcYH6DXLpYomzX+hPf 6YvC8sVL41pQM29WJb7ld4W3+9jKEgydGBTbjtTteD0xR2N1uGMs0cX1sQyJYz4QdqJRnCBZOmq IfIaiUYTo2MpUZGwQDEklhAY1pFnf6nvApPWjMs8NHqmCMS5XAoXaV X-Received: by 2002:a05:6000:2410:b0:435:a2f8:1536 with SMTP id ffacd0b85a97d-435dd1c0ca3mr8693194f8f.43.1769622754829; Wed, 28 Jan 2026 09:52:34 -0800 (PST) Received: from phoenix.local (204-195-96-226.wavecable.com. [204.195.96.226]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-435e10e476dsm8898944f8f.4.2026.01.28.09.52.33 for (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 28 Jan 2026 09:52:34 -0800 (PST) Message-ID: <697a4ce2.df0a0220.a9380.da00@mx.google.com> Date: Wed, 28 Jan 2026 09:52:34 -0800 (PST) From: Stephen Hemminger To: dev@dpdk.org Subject: [REVIEW] common/cnxk: fix duplicate branch compiler warning In-Reply-To: <20260128060316.3967610-18-rbhansali@marvell.com> References: <20260128060316.3967610-18-rbhansali@marvell.com> Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org AI-generated review of bundle-1704-common-cnxk.mbox Reviewed using Claude (claude-opus-4-5-20251101) This is an automated review. Please verify all suggestions. --- # DPDK Patch Review: bundle-1704-common-cnxk.mbox ## Overview This is a patch series of 18 patches for DPDK common/cnxk and net/cnxk driver= s, primarily dealing with inline IPsec improvements for CN20K platform. --- ## Patch 1/18: common/cnxk: fix engine capabilities fetch logic ### Commit Message **Warning**: The commit message body lacks detail. The description "This patc= h fixes the same" is vague and should explain what the actual fix is. **Info**: The `Fixes:` tag is correctly formatted with 12-character SHA. ### Code Review **Error**: In `cnxk_security.c`, lines 1400 and 1409 - The change from `ROC_I= E_OT_SA_LIFE_UNIT_*` to `ROC_IE_OW_SA_LIFE_UNIT_*` macros appears correct for= the OW (cn20k) context. **Error**: In `roc_nix_inl.c`, line 1231 - The loop condition change from `wh= ile (lmt_status !=3D 0)` to `while (lmt_status =3D=3D 0)` is a significant lo= gic change. The commit message should explain why the previous condition was = wrong (it was causing an infinite loop as stated). --- ## Patch 2/18: common/cnxk: remove dependency on cryptodev for RXC ### Commit Message **Info**: Good descriptive subject line. ### Code Review **Info**: The code correctly moves the cryptodev check inside the `roc_model_= is_cn10k()` block, allowing cn20k to proceed without this dependency. --- ## Patch 3/18: common/cnxk: support inbound pdb configuration ### Commit Message **Info**: Subject and body are acceptable. ### Code Review **Info**: Simple addition of `pdb_ena` field usage in mbox configuration. --- ## Patch 4/18: common/cnxk: update CPT RXC structures ### Commit Message **Warning**: The subject says "update" but should be more specific about what= was updated (byte order/endianness handling). ### Code Review **Info**: The structures are being reorganized with proper unions for differe= nt platforms (cn10k vs generic). The byte order appears to be reversed in the= new `cpt_frag_info_s` compared to `cpt_cn10k_frag_info_s`. --- ## Patch 5/18: common/cnxk: update inline profile ID for cn20k ### Commit Message **Info**: Acceptable. ### Code Review **Info**: Correctly adds cn20k-specific profile ID handling. --- ## Patch 6/18: common/cnxk: update inline RQ mask ### Commit Message **Info**: Acceptable. ### Code Review **Warning**: Significant code removal - the `nix_inl_rq_mask_cfg` function re= moves the SPB setup via `nix_rx_inl_lf_cfg` mbox call. This should be validat= ed that it's not needed for cn20k. --- ## Patch 7/18: net/cnxk: avoid security flag for custom inbound SA ### Commit Message **Info**: Good descriptive subject. ### Code Review **Info**: Simple conditional change that looks correct. --- ## Patch 8/18: net/cnxk: add CPT code check for soft expiry ### Commit Message **Info**: Acceptable. ### Code Review **Info**: Adding `ROC_IE_OW_UCC_SUCCESS_SA_SOFTEXP_AGAIN` case is a valid add= ition. --- ## Patch 9/18: net/cnxk: skip write SA for cn20k ### Commit Message **Info**: Acceptable but could explain why write SA should be skipped. ### Code Review **Info**: Simple flag change. --- ## Patch 10/18: net/cnxk: update NIX reassembly fast path ### Commit Message **Info**: Acceptable. ### Code Review **Warning**: The function `nix_cqe_xtract_mseg` now has an `#if defined(RTE_A= RCH_ARM64)` guard with an empty stub for non-ARM64. This may break functional= ity on other architectures silently. **Info**: The use of NEON intrinsics (`vreinterpret_u16_u64`, `vrev64_u16`, e= tc.) is appropriate for ARM64 optimization. --- ## Patch 11/18: net/cnxk: update aura batch free ### Commit Message **Info**: Subject could be more descriptive about what was updated (mask chan= ge from 0x1 to 0x3). ### Code Review **Info**: The mask change from `0x1` to `0x3` appears consistent across all o= ccurrences. --- ## Patch 12/18: net/cnxk: update fastpath function for OOP ### Commit Message **Info**: OOP should be expanded (Out-Of-Place) in the commit body for clarit= y. ### Code Review **Info**: Adds OOP (out-of-place) processing support. The `nix_sec_oop_proces= s` function is well-implemented. **Info**: The function correctly handles the alternate encrypted-decrypted po= inter pattern in the gather list. --- ## Patch 13/18: event/cnxk: update fastpath function for OOP ### Commit Message **Info**: Acceptable. ### Code Review **Info**: Consistent OOP handling with the net/cnxk changes. --- ## Patch 14/18: common/cnxk: flow rule config for non-inplace ### Commit Message **Info**: Acceptable. ### Code Review **Info**: Adds `is_non_inp` flag handling in flow rules. --- ## Patch 15/18: net/cnxk: enable PDB in IPsec outbound path ### Commit Message **Info**: PDB should be expanded in the commit body. ### Code Review **Info**: Simple bit flag additions for PDB enablement. --- ## Patch 16/18: common/cnxk: initialize CPT LF for CQ config ### Commit Message **Info**: Acceptable. ### Code Review **Info**: Properly initializes CPT LF for CQ configurations with appropriate = cleanup in the error path and fini function. --- ## Patch 17/18: common/cnxk: fix CPT CQ roll over handling ### Commit Message **Info**: Good - includes proper `Fixes:` tag. ### Code Review **Info**: The fix correctly handles ring rollover by tracking `cq_head` separ= ately and using modulo arithmetic. **Warning**: Line 102-103 adds error logging with `plt_err` but this could be= noisy in production. Consider using `plt_warn` or adding rate limiting. --- ## Patch 18/18: common/cnxk: fix duplicate branch compiler warning ### Commit Message **Info**: Good - includes Bugzilla ID and proper `Fixes:` tags. ### Code Review **Info**: The `PLT_STATIC_ASSERT` additions verify that SA sizes are equal ac= ross platforms, which justifies using a common macro. **Warning**: In `roc_nix_priv.h`, lines 26 and 29 - changing integer division= to floating point division (`100` to `100.0`) changes the type of the result= . Verify this doesn't cause issues with consumers of these macros. ```c #define NIX_CQ_LBP_THRESH_FRAC_REF1 (80 * 16 / 100.0) #define NIX_CQ_LBP_THRESH_FRAC (75 * 16 / 100.0) ``` --- ## General Issues Across All Patches ### License Compliance **Info**: All modified files appear to have proper SPDX license identifiers a= lready present. ### Missing Items **Warning**: No release notes updates found in this patch series despite API/= behavior changes. Consider adding entries to `doc/guides/rel_notes/`. **Warning**: Several patches lack `Cc: stable@dpdk.org` tag despite being bug= fixes (patches 1, 17, 18). ### Coding Style **Info**: Code generally follows DPDK coding standards with proper indentatio= n and naming conventions. --- ## Summary | Severity | Count | |----------|-------| | Error | 0 | | Warning | 8 | | Info | Multiple | The patch series is generally well-structured. Main concerns are: 1. Missing `Cc: stable@dpdk.org` on fix patches 2. Missing release notes 3. Some commit messages could be more descriptive 4. Floating point division change in macros should be verified