<!DOCTYPE html><html><head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body> Hi Sunyuechi, <div class="moz-cite-prefix">On 04/06/2025 12:39, 孙越池 wrote: </div> <blockquote type="cite" cite="mid:26aab158.28484.1973abd6624.Coremail.sunyuechi@iscas.ac.cn"> > why is it done in a scalar way instead of using `__riscv_vsrl_vx_u32m1()?` I assume you're relying on the compiler here? I don't know the exact reason, but based on experience, using indexed loads tends to be slower for small-scale and low-computation cases. So I've tried both methods. In this case, if using `vsrl`, it would require `__riscv_vluxei32_v_u32m1`, which is much slower. ``` vuint32m1_t vip_shifted = __riscv_vsll_vx_u32m1(__riscv_vsrl_vx_u32m1(__riscv_vle32_v_u32m1((const uint32_t *)&ip, vl), 8, vl), 2, vl); vuint32m1_t vtbl_entry = __riscv_vluxei32_v_u32m1( (const uint32_t *)(lpm->tbl24), vip_shifted, vl); ``` > have you redefined the xmm_t type for proper index addressing? It is in `eal/riscv/include/rte_vect.h:` ``` typedef int32_t xmm_t __attribute__((vector_size(16))); ``` > I'd recommend that you use FIB to select an implementation at runtime. All the rest LPM vector x4 implementations are done this way, and their code is inlined. > Also, please consider writing a slightly more informative and explanatory commit message. </blockquote> The commit message still looks uninformative to me: >lpm_perf_autotest on BPI-F3 we have no idea what's that > scalar: 5.7 cycles I'm not sure we want to have this information in commit message as well, because it is useless. Cycles depends on so much variable parts - what freq of the CPU was, what speed of memory, size of caches, and so on. This information is irrelevant and become obsolete pretty fast. From the latest commit: >The best way ... However, ... Therefore, ... this commit does not modify >Unifying the code style between lpm and fib may be worth considering in the future. I don't think this is a good idea to put into the commit message information about what was NOT done. You should put all this information (platform you were running, performance, implementation considerations and thoughts) into the patch notes. <blockquote type="cite" cite="mid:26aab158.28484.1973abd6624.Coremail.sunyuechi@iscas.ac.cn"> I agree that the FIB approach is clearly better here, but adopting this method would require changing the function initialization logic for all architectures in LPM, as well as updating the relevant structures. I'm not sure it's worth doing right now, since this commit is intended to be just a small change for RISC-V. I'm more inclined to follow the existing structure and avoid touching other architectures' code. Would it be acceptable to leave this kind of refactoring for the future? If you're certain it should be done now, I'll make the changes. For now, I've only updated the commit message to include this idea (v2). <div style="white-space:nowrap;"> </div> </blockquote> I'm not talking about adopting the FIB approach to the LPM. Instead, I suggested keeping LPM code consistent and leaving your implementation as a static inline function. And if you want to have runtime CPU flags check - you're welcome to do so in the FIB. <blockquote type="cite" cite="mid:26aab158.28484.1973abd6624.Coremail.sunyuechi@iscas.ac.cn"> <blockquote name="replyContent" class="ReferenceQuote" style="font-family:SimSun;padding-left:5px;margin-left:5px;border-left:2px solid #B6B6B6;margin-right:0px;"> -----原始邮件----- 发件人:"Medvedkin, Vladimir" <a class="moz-txt-link-rfc2396E" href="mailto:vladimir.medvedkin@intel.com"><vladimir.medvedkin@intel.com></a> 发送时间:2025-05-30 21:13:57 (星期五) 收件人: <a class="moz-txt-link-abbreviated" href="mailto:uk7b@foxmail.com">uk7b@foxmail.com</a>, <a class="moz-txt-link-abbreviated" href="mailto:dev@dpdk.org">dev@dpdk.org</a> 抄送: sunyuechi <a class="moz-txt-link-rfc2396E" href="mailto:sunyuechi@iscas.ac.cn"><sunyuechi@iscas.ac.cn></a>, "Thomas Monjalon" <a class="moz-txt-link-rfc2396E" href="mailto:thomas@monjalon.net"><thomas@monjalon.net></a>, "Bruce Richardson" <a class="moz-txt-link-rfc2396E" href="mailto:bruce.richardson@intel.com"><bruce.richardson@intel.com></a>, "Stanislaw Kardach" <a class="moz-txt-link-rfc2396E" href="mailto:stanislaw.kardach@gmail.com"><stanislaw.kardach@gmail.com></a> 主题: Re: [PATCH 2/3] lib/lpm: R-V V rte_lpm_lookupx4 Hi c, <div class="moz-cite-prefix"> On 28/05/2025 18:00, <a class="moz-txt-link-abbreviated moz-txt-link-freetext" href="mailto:uk7b@foxmail.com" moz-do-not-send="true">uk7b@foxmail.com</a> wrote: </div> <blockquote type="cite" cite="mid:tencent_A2BD85C9F3B0AA4658640DA91443CDF5630A@qq.com"> <pre wrap="" class="moz-quote-pre">From: sunyuechi <a class="moz-txt-link-rfc2396E" href="mailto:sunyuechi@iscas.ac.cn" moz-do-not-send="true"><sunyuechi@iscas.ac.cn></a> bpi-f3: scalar: 5.7 cycles rvv: 2.4 cycles Maybe runtime detection in LPM should be added for all architectures, but this commit is only about the RVV part. </pre> </blockquote> I would advise you to look into the FIB library, it has exactly what you are looking for. Also, please consider writing a slightly more informative and explanatory commit message. <blockquote type="cite" cite="mid:tencent_A2BD85C9F3B0AA4658640DA91443CDF5630A@qq.com"> <pre wrap="" class="moz-quote-pre">Signed-off-by: sunyuechi <a class="moz-txt-link-rfc2396E" href="mailto:sunyuechi@iscas.ac.cn" moz-do-not-send="true"><sunyuechi@iscas.ac.cn></a> --- MAINTAINERS | 2 + lib/lpm/meson.build | 1 + lib/lpm/rte_lpm.h | 2 + lib/lpm/rte_lpm_rvv.h | 91 +++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 96 insertions(+) create mode 100644 lib/lpm/rte_lpm_rvv.h </pre> </blockquote> <snip> <blockquote type="cite" cite="mid:tencent_A2BD85C9F3B0AA4658640DA91443CDF5630A@qq.com"> <pre wrap="" class="moz-quote-pre">+static inline void rte_lpm_lookupx4_rvv( + const struct rte_lpm *lpm, xmm_t ip, uint32_t hop[4], uint32_t defv) +{ + size_t vl = 4; + + const uint32_t *tbl24_p = (const uint32_t *)lpm->tbl24; + uint32_t tbl_entries[4] = { + tbl24_p[((uint32_t)ip[0]) >> 8], + tbl24_p[((uint32_t)ip[1]) >> 8], + tbl24_p[((uint32_t)ip[2]) >> 8], + tbl24_p[((uint32_t)ip[3]) >> 8], + };</pre> </blockquote> I'm not an expert in RISC-V, but why is it done in a scalar way instead of using __riscv_vsrl_vx_u32m1()? I assume you're relying on the compiler here? Also, have you redefined the xmm_t type for proper index addressing? <blockquote type="cite" cite="mid:tencent_A2BD85C9F3B0AA4658640DA91443CDF5630A@qq.com"> <pre wrap="" class="moz-quote-pre">+ vuint32m1_t vtbl_entry = __riscv_vle32_v_u32m1(tbl_entries, vl); + + vbool32_t mask = __riscv_vmseq_vx_u32m1_b32( + __riscv_vand_vx_u32m1(vtbl_entry, RTE_LPM_VALID_EXT_ENTRY_BITMASK, vl), + RTE_LPM_VALID_EXT_ENTRY_BITMASK, vl); </pre> </blockquote> <snip> <blockquote type="cite" cite="mid:tencent_A2BD85C9F3B0AA4658640DA91443CDF5630A@qq.com"> <pre wrap="" class="moz-quote-pre">+ +static inline void rte_lpm_lookupx4( + const struct rte_lpm *lpm, xmm_t ip, uint32_t hop[4], uint32_t defv) +{ + lpm_lookupx4_impl(lpm, ip, hop, defv); +} + +RTE_INIT(rte_lpm_init_alg) +{ + lpm_lookupx4_impl = rte_cpu_get_flag_enabled(RTE_CPUFLAG_RISCV_ISA_V) + ? rte_lpm_lookupx4_rvv + : rte_lpm_lookupx4_scalar; +}</pre> </blockquote> As I mentioned earlier, I'd recommend that you use FIB to select an implementation at runtime. All the rest LPM vector x4 implementations are done this way, and their code is inlined. <blockquote type="cite" cite="mid:tencent_A2BD85C9F3B0AA4658640DA91443CDF5630A@qq.com"> <pre wrap="" class="moz-quote-pre">+ +#ifdef __cplusplus +} +#endif + +#endif /* _RTE_LPM_RVV_H_ */ </pre> </blockquote> <pre class="moz-signature" cols="72">-- Regards, Vladimir</pre> </blockquote> </blockquote> <pre class="moz-signature" cols="72">-- Regards, Vladimir</pre> </body> </html>