From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 4480DC3DA7D for ; Tue, 3 Jan 2023 14:12:56 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237840AbjACOMy (ORCPT ); Tue, 3 Jan 2023 09:12:54 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58924 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237837AbjACOMl (ORCPT ); Tue, 3 Jan 2023 09:12:41 -0500 Received: from mail-pj1-x102e.google.com (mail-pj1-x102e.google.com [IPv6:2607:f8b0:4864:20::102e]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id C312911A13 for ; Tue, 3 Jan 2023 06:12:39 -0800 (PST) Received: by mail-pj1-x102e.google.com with SMTP id o21so2861744pjw.0 for ; Tue, 03 Jan 2023 06:12:39 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=F7ob+2vX2/dVNm5GFakcAOJyyRuKnpkny5ZMoNumsbg=; b=SC2efmYBg1+mzWG7pzPU4iMr+PFRDzvoSWpPlMH6YKZt+hR0iPxqcmZWSNcJZCwK+V wGRjiEBBWXFHOjS8klcs14FF4xiO7xLK0pCoqBYolvIG1iykE4ltt2whCD6AO6Aso8Kn MzrWZDqM22mMPuAXNi7yVnIy/CsR5G1o0diamtaLsY1wKt4FJ+iavxU0hkGBvYYRtGxP +LGdmSucXk4gNVRAfh678TYbLDK0rOVPEpKdlx1zTtJ/AqLzEk8LgJ9q+0h9Lg16rfzH aHkp/bUkC7wYGRYNge/xd9Kd/a/t0IGbIYnEGcKqq8VqYT6t2dXt6RNKDrkctboafOl8 TwTw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=F7ob+2vX2/dVNm5GFakcAOJyyRuKnpkny5ZMoNumsbg=; b=n9VRzfAKkPbE3ZphuKazUwR8bwQGNjkQ6atj5xAzlUAZV0lTL0wJSSF9XM/Sj0dpUJ hRKmyXTHhnoObq3JLG5AP60YEAHVfQv/WFXsUToR4MdD+Y5/yhtwYaHkQpmVY5et27gj DhO5FDpltI/DRmSWYLgPgTJvNCXX6VcE7oBLTozndnH1gEB7Bhm3fkCx67XwU9vhdnnp jLUmtsCFk0dNu93XRpYRNm4l6Gj5fn91ItRWtkXLUxTr3zIxbeCPo7xZbpudAxHx5iqb msDs2lXGB/yF3OtOffxRSEOrjxA0ZHlHwby1MbvPSU6htwdtBGibnt0b4kNLiB3dklB1 S71w== X-Gm-Message-State: AFqh2kqkumUBHDfRQTzzevG2KZ/0O7xyDjZMOjvLYK+4tFkTybxoYswm s9efsuyLtpTm7ka4cv6mtVswmA== X-Google-Smtp-Source: AMrXdXsdaNFT9PfPhIcXX2+sh5QFHWfKnommCCIjPlGRE8sR3Zqyk6aDdAW9OIFDQIqGWaWuBKBr1g== X-Received: by 2002:a17:903:2411:b0:192:d9dd:167d with SMTP id e17-20020a170903241100b00192d9dd167dmr3077913plo.43.1672755159141; Tue, 03 Jan 2023 06:12:39 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.12.33 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:12:38 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel , Bin Meng , Atish Patra Subject: [PATCH v16 1/9] RISC-V: Clear SIP bit only when using SBI IPI operations Date: Tue, 3 Jan 2023 19:42:13 +0530 Message-Id: <20230103141221.772261-2-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The software interrupt pending (i.e. [M|S]SIP) bit is writeable for S-mode but read-only for M-mode so we clear this bit only when using SBI IPI operations. Signed-off-by: Anup Patel Reviewed-by: Bin Meng Reviewed-by: Atish Patra Acked-by: Palmer Dabbelt --- arch/riscv/kernel/sbi.c | 8 +++++++- arch/riscv/kernel/smp.c | 2 -- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c index 5c87db8fdff2..ac99a70ead6a 100644 --- a/arch/riscv/kernel/sbi.c +++ b/arch/riscv/kernel/sbi.c @@ -646,8 +646,14 @@ static void sbi_send_cpumask_ipi(const struct cpumask = *target) sbi_send_ipi(target); } =20 +static void sbi_ipi_clear(void) +{ + csr_clear(CSR_IP, IE_SIE); +} + static const struct riscv_ipi_ops sbi_ipi_ops =3D { - .ipi_inject =3D sbi_send_cpumask_ipi + .ipi_inject =3D sbi_send_cpumask_ipi, + .ipi_clear =3D sbi_ipi_clear }; =20 void __init sbi_init(void) diff --git a/arch/riscv/kernel/smp.c b/arch/riscv/kernel/smp.c index 8c3b59f1f9b8..8a12768c09ee 100644 --- a/arch/riscv/kernel/smp.c +++ b/arch/riscv/kernel/smp.c @@ -112,8 +112,6 @@ void riscv_clear_ipi(void) { if (ipi_ops && ipi_ops->ipi_clear) ipi_ops->ipi_clear(); - - csr_clear(CSR_IP, IE_SIE); } EXPORT_SYMBOL_GPL(riscv_clear_ipi); =20 --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id A1AC4C3DA7D for ; Tue, 3 Jan 2023 14:13:07 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237811AbjACONG (ORCPT ); Tue, 3 Jan 2023 09:13:06 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59240 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237818AbjACOMq (ORCPT ); Tue, 3 Jan 2023 09:12:46 -0500 Received: from mail-pj1-x1032.google.com (mail-pj1-x1032.google.com [IPv6:2607:f8b0:4864:20::1032]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id D43721117F for ; Tue, 3 Jan 2023 06:12:44 -0800 (PST) Received: by mail-pj1-x1032.google.com with SMTP id w4-20020a17090ac98400b002186f5d7a4cso36041857pjt.0 for ; Tue, 03 Jan 2023 06:12:44 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=GBh16wlKK1pVBXMjv15TuW7j92X6+xaNPp74Ke5cwzg=; b=ag2VYZdi8sHLadDfTOWomK5QgMIp5UxmtALzG6C+VynD9m/7Y1X1UTRRMNa1SLWudR fsKqlD5togrJ6kiHvfsCqE121VmJzDJS4OvJpZT49IeV1uoNPMuy3lPB7mmhACLJ377C sSnhYyWZN3HBMbt3l4rlQWsYKweAVSL+AgdyQr6/jobS5f64v+Y/jJTHZopbYCZmGExD +ITdd8Wp5MEOBw17CkhiQDkBEdt9hL6zz/QfFan4eCTh4ufZMW+SEW+RirbywdY/cLgc mDZt0aeayvg/vFK0p78cMcIs7WgabCpHW+aYgYlFKLMgT2FvmzXhuQgPmfHbNlXLKxTp Voig== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=GBh16wlKK1pVBXMjv15TuW7j92X6+xaNPp74Ke5cwzg=; b=zB+x7xGBor9Rm/2cxoF7/wdzGUe4xE1llNp9WMQ6EBTzpTklSs+5jEOol6cdiQT/xP DEMPixvdY7y0T84gyvW2AkBRK63r/FbTUDGmXlCmJqsy5TT2jYYfH7VDJd6lUpmyzvFk 7bP4VEsogm8qoEN+2cozRdVmezaTPRUzDBXZqLVgiUpls31u48er+bXnDcXWUFAVe6pm PomXWmqmNmmYO6HdHHyoUNXgi0jN5Xxtmpbk7N2VBsTWMfL79kuMUfwbaDMdPxFT2ZBp hoHAqAwJ3qxFEPpjR1+0e2p2NLO8Rz3ZA3I+wXfcOpC87FiA6IFoWIRhELEdFgt8y6T2 as7w== X-Gm-Message-State: AFqh2kr0gUatTPjUbGIRiZO5Yx+VTfuASDKlOgpzkHGSEWNgozRmbCao 0Z103/UuXW9o3+6dMfx2XbdCVQ== X-Google-Smtp-Source: AMrXdXuLqYi4hbftZX6Ald4rvwIshOeG5B3xI5B05B6I0NSM8AMUySmbHh3I2kBz7X9fDRWmiUfGUg== X-Received: by 2002:a17:902:b20b:b0:186:7a6b:24d9 with SMTP id t11-20020a170902b20b00b001867a6b24d9mr50119651plr.23.1672755164157; Tue, 03 Jan 2023 06:12:44 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.12.39 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:12:43 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel , Atish Patra Subject: [PATCH v16 2/9] irqchip/riscv-intc: Allow drivers to directly discover INTC hwnode Date: Tue, 3 Jan 2023 19:42:14 +0530 Message-Id: <20230103141221.772261-3-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Various RISC-V drivers (such as SBI IPI, SBI Timer, SBI PMU, and KVM RISC-V) don't have associated DT node but these drivers need standard per-CPU (local) interrupts defined by the RISC-V privileged specification. We add riscv_get_intc_hwnode() in arch/riscv which allows RISC-V drivers not having DT node to discover INTC hwnode which in-turn helps these drivers to map per-CPU (local) interrupts provided by the INTC driver. Signed-off-by: Anup Patel Reviewed-by: Atish Patra Acked-by: Palmer Dabbelt --- arch/riscv/include/asm/irq.h | 4 ++++ arch/riscv/kernel/irq.c | 18 ++++++++++++++++++ drivers/irqchip/irq-riscv-intc.c | 7 +++++++ 3 files changed, 29 insertions(+) diff --git a/arch/riscv/include/asm/irq.h b/arch/riscv/include/asm/irq.h index e4c435509983..43b9ebfbd943 100644 --- a/arch/riscv/include/asm/irq.h +++ b/arch/riscv/include/asm/irq.h @@ -12,6 +12,10 @@ =20 #include =20 +void riscv_set_intc_hwnode_fn(struct fwnode_handle *(*fn)(void)); + +struct fwnode_handle *riscv_get_intc_hwnode(void); + extern void __init init_IRQ(void); =20 #endif /* _ASM_RISCV_IRQ_H */ diff --git a/arch/riscv/kernel/irq.c b/arch/riscv/kernel/irq.c index 7207fa08d78f..96d3171f0ca1 100644 --- a/arch/riscv/kernel/irq.c +++ b/arch/riscv/kernel/irq.c @@ -7,9 +7,27 @@ =20 #include #include +#include +#include #include #include =20 +static struct fwnode_handle *(*__get_intc_node)(void); + +void riscv_set_intc_hwnode_fn(struct fwnode_handle *(*fn)(void)) +{ + __get_intc_node =3D fn; +} + +struct fwnode_handle *riscv_get_intc_hwnode(void) +{ + if (__get_intc_node) + return __get_intc_node(); + + return NULL; +} +EXPORT_SYMBOL_GPL(riscv_get_intc_hwnode); + int arch_show_interrupts(struct seq_file *p, int prec) { show_ipi_stats(p, prec); diff --git a/drivers/irqchip/irq-riscv-intc.c b/drivers/irqchip/irq-riscv-i= ntc.c index 499e5f81b3fe..9066467e99e4 100644 --- a/drivers/irqchip/irq-riscv-intc.c +++ b/drivers/irqchip/irq-riscv-intc.c @@ -92,6 +92,11 @@ static const struct irq_domain_ops riscv_intc_domain_ops= =3D { .xlate =3D irq_domain_xlate_onecell, }; =20 +static struct fwnode_handle *riscv_intc_hwnode(void) +{ + return intc_domain->fwnode; +} + static int __init riscv_intc_init(struct device_node *node, struct device_node *parent) { @@ -126,6 +131,8 @@ static int __init riscv_intc_init(struct device_node *n= ode, return rc; } =20 + riscv_set_intc_hwnode_fn(riscv_intc_hwnode); + cpuhp_setup_state(CPUHP_AP_IRQ_RISCV_STARTING, "irqchip/riscv/intc:starting", riscv_intc_cpu_starting, --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 04E0CC3DA7D for ; Tue, 3 Jan 2023 14:13:12 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237842AbjACONJ (ORCPT ); Tue, 3 Jan 2023 09:13:09 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58492 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237834AbjACOMv (ORCPT ); Tue, 3 Jan 2023 09:12:51 -0500 Received: from mail-pj1-x102c.google.com (mail-pj1-x102c.google.com [IPv6:2607:f8b0:4864:20::102c]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 0B07A11445 for ; Tue, 3 Jan 2023 06:12:50 -0800 (PST) Received: by mail-pj1-x102c.google.com with SMTP id v13-20020a17090a6b0d00b00219c3be9830so31186303pjj.4 for ; Tue, 03 Jan 2023 06:12:50 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=VeIojzaABUKnNOf6/IOzgNVZ1YxXJw15pDsAxwDwZ7k=; b=jWjUf0YdK2SMD8Hh2MCAwUrCm2W91Xxt3BfSJbQiiG0+KxAPqJAL6/IR8XlkKfbu2f lj3ETdyHU9TGn4KNnratw7EC+UiiqAJA2/mTAGHD/D46nkvSyQpJcL0IUCZYb2AbwqYH SeDNg3XXMMXKi23uVKNzQys4jGETFgmDZOFcAr14JrCger9B33h1F3olA6epPMn9V+bQ XpgYmUYiQ/V5RtcA65upxEB0J7ehM+sY08CnaxJvwppItpNrjN90US7osagz3XFFhq83 zJesW5bPfxSPXikF9w9ylOEfI+mk8v4NZa7NNjEZ47EpLMbj7G585FuHhlPRfqJSxI0s +6yA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=VeIojzaABUKnNOf6/IOzgNVZ1YxXJw15pDsAxwDwZ7k=; b=tQ3pPP/jHO+pRWwTasOy51avGyPQHXxz7z4w/c4NAGFius0d8NOJXqSOOWORm55lAs L28XnvunK0ZrXzkAoh4Xpezc8Ho++YZP3AbMfAAuqqUqVeY7K5athZr+5dR6I18OLgae jlhcanvXGRdHMEZBEVr+0t2YWlI6/RioOwnxI8VrVhMlRnHU2PgpsCrwITsS/qxHChJS g6bX1xDjVO8kVfUQ1auA8wY9A6PgLBV27QsLWzGLtf/G2u8QJnoXsj8Q3dTTTahJjkpU O50IU5TTORfh55SzUB8/SCk0LUuJC3MchW1x68U5WvUfr7/6mUeGQBjdJ3nFnvyFkQUO WZ5A== X-Gm-Message-State: AFqh2kpTj4VOh3rL84O7JnMY8tdSPi2UhRwI/PvskVetnonrQhmCjtVD uVR/IRHRjYbOaq38eaYB5ON5vQ== X-Google-Smtp-Source: AMrXdXsbQVf67b73TY05kz+WINqDSvAUcG7y9AHvgmKqoJEbL6J2YbDn5N4Zb1KZgK8LsrKyiZz3Ig== X-Received: by 2002:a17:903:3287:b0:192:e0ab:a4a2 with SMTP id jh7-20020a170903328700b00192e0aba4a2mr1227619plb.51.1672755169260; Tue, 03 Jan 2023 06:12:49 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.12.44 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:12:48 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel Subject: [PATCH v16 3/9] genirq: Add mechanism to multiplex a single HW IPI Date: Tue, 3 Jan 2023 19:42:15 +0530 Message-Id: <20230103141221.772261-4-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" All RISC-V platforms have a single HW IPI provided by the INTC local interrupt controller. The HW method to trigger INTC IPI can be through external irqchip (e.g. RISC-V AIA), through platform specific device (e.g. SiFive CLINT timer), or through firmware (e.g. SBI IPI call). To support multiple IPIs on RISC-V, add a generic IPI multiplexing mechanism which help us create multiple virtual IPIs using a single HW IPI. This generic IPI multiplexing is inspired by the Apple AIC irqchip driver and it is shared by various RISC-V irqchip drivers. Signed-off-by: Anup Patel Reviewed-by: Hector Martin Tested-by: Hector Martin Acked-by: Palmer Dabbelt --- include/linux/irq.h | 3 + kernel/irq/Kconfig | 5 ++ kernel/irq/Makefile | 1 + kernel/irq/ipi-mux.c | 207 +++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 216 insertions(+) create mode 100644 kernel/irq/ipi-mux.c diff --git a/include/linux/irq.h b/include/linux/irq.h index c3eb89606c2b..b1b28affb32a 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -1266,6 +1266,9 @@ int __ipi_send_mask(struct irq_desc *desc, const stru= ct cpumask *dest); int ipi_send_single(unsigned int virq, unsigned int cpu); int ipi_send_mask(unsigned int virq, const struct cpumask *dest); =20 +void ipi_mux_process(void); +int ipi_mux_create(unsigned int nr_ipi, void (*mux_send)(unsigned int cpu)= ); + #ifdef CONFIG_GENERIC_IRQ_MULTI_HANDLER /* * Registers a generic IRQ handling function as the top-level IRQ handler = in diff --git a/kernel/irq/Kconfig b/kernel/irq/Kconfig index b64c44ae4c25..2531f3496ab6 100644 --- a/kernel/irq/Kconfig +++ b/kernel/irq/Kconfig @@ -86,6 +86,11 @@ config GENERIC_IRQ_IPI depends on SMP select IRQ_DOMAIN_HIERARCHY =20 +# Generic IRQ IPI Mux support +config GENERIC_IRQ_IPI_MUX + bool + depends on SMP + # Generic MSI hierarchical interrupt domain support config GENERIC_MSI_IRQ bool diff --git a/kernel/irq/Makefile b/kernel/irq/Makefile index b4f53717d143..f19d3080bf11 100644 --- a/kernel/irq/Makefile +++ b/kernel/irq/Makefile @@ -15,6 +15,7 @@ obj-$(CONFIG_GENERIC_IRQ_MIGRATION) +=3D cpuhotplug.o obj-$(CONFIG_PM_SLEEP) +=3D pm.o obj-$(CONFIG_GENERIC_MSI_IRQ) +=3D msi.o obj-$(CONFIG_GENERIC_IRQ_IPI) +=3D ipi.o +obj-$(CONFIG_GENERIC_IRQ_IPI_MUX) +=3D ipi-mux.o obj-$(CONFIG_SMP) +=3D affinity.o obj-$(CONFIG_GENERIC_IRQ_DEBUGFS) +=3D debugfs.o obj-$(CONFIG_GENERIC_IRQ_MATRIX_ALLOCATOR) +=3D matrix.o diff --git a/kernel/irq/ipi-mux.c b/kernel/irq/ipi-mux.c new file mode 100644 index 000000000000..3a403c3a785d --- /dev/null +++ b/kernel/irq/ipi-mux.c @@ -0,0 +1,207 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Multiplex several virtual IPIs over a single HW IPI. + * + * Copyright The Asahi Linux Contributors + * Copyright (c) 2022 Ventana Micro Systems Inc. + */ + +#define pr_fmt(fmt) "ipi-mux: " fmt +#include +#include +#include +#include +#include +#include +#include +#include +#include + +struct ipi_mux_cpu { + atomic_t enable; + atomic_t bits; +}; + +static struct ipi_mux_cpu __percpu *ipi_mux_pcpu; +static struct irq_domain *ipi_mux_domain; +static void (*ipi_mux_send)(unsigned int cpu); + +static void ipi_mux_mask(struct irq_data *d) +{ + struct ipi_mux_cpu *icpu =3D this_cpu_ptr(ipi_mux_pcpu); + + atomic_andnot(BIT(irqd_to_hwirq(d)), &icpu->enable); +} + +static void ipi_mux_unmask(struct irq_data *d) +{ + struct ipi_mux_cpu *icpu =3D this_cpu_ptr(ipi_mux_pcpu); + u32 ibit =3D BIT(irqd_to_hwirq(d)); + + atomic_or(ibit, &icpu->enable); + + /* + * The atomic_or() above must complete before the atomic_read() + * below to avoid racing ipi_mux_send_mask(). + */ + smp_mb__after_atomic(); + + /* If a pending IPI was unmasked, raise a parent IPI immediately. */ + if (atomic_read(&icpu->bits) & ibit) + ipi_mux_send(smp_processor_id()); +} + +static void ipi_mux_send_mask(struct irq_data *d, const struct cpumask *ma= sk) +{ + struct ipi_mux_cpu *icpu =3D this_cpu_ptr(ipi_mux_pcpu); + u32 ibit =3D BIT(irqd_to_hwirq(d)); + unsigned long pending; + int cpu; + + for_each_cpu(cpu, mask) { + icpu =3D per_cpu_ptr(ipi_mux_pcpu, cpu); + + /* + * This sequence is the mirror of the one in ipi_mux_unmask(); + * see the comment there. Additionally, release semantics + * ensure that the vIPI flag set is ordered after any shared + * memory accesses that precede it. This therefore also pairs + * with the atomic_fetch_andnot in ipi_mux_process(). + */ + pending =3D atomic_fetch_or_release(ibit, &icpu->bits); + + /* + * The atomic_fetch_or_release() above must complete + * before the atomic_read() below to avoid racing with + * ipi_mux_unmask(). + */ + smp_mb__after_atomic(); + + /* + * The flag writes must complete before the physical IPI is + * issued to another CPU. This is implied by the control + * dependency on the result of atomic_read() below, which is + * itself already ordered after the vIPI flag write. + */ + if (!(pending & ibit) && (atomic_read(&icpu->enable) & ibit)) + ipi_mux_send(cpu); + } +} + +static const struct irq_chip ipi_mux_chip =3D { + .name =3D "IPI Mux", + .irq_mask =3D ipi_mux_mask, + .irq_unmask =3D ipi_mux_unmask, + .ipi_send_mask =3D ipi_mux_send_mask, +}; + +static int ipi_mux_domain_alloc(struct irq_domain *d, unsigned int virq, + unsigned int nr_irqs, void *arg) +{ + int i; + + for (i =3D 0; i < nr_irqs; i++) { + irq_set_percpu_devid(virq + i); + irq_domain_set_info(d, virq + i, i, &ipi_mux_chip, NULL, + handle_percpu_devid_irq, NULL, NULL); + } + + return 0; +} + +static const struct irq_domain_ops ipi_mux_domain_ops =3D { + .alloc =3D ipi_mux_domain_alloc, + .free =3D irq_domain_free_irqs_top, +}; + +/** + * ipi_mux_process - Process multiplexed virtual IPIs + */ +void ipi_mux_process(void) +{ + struct ipi_mux_cpu *icpu =3D this_cpu_ptr(ipi_mux_pcpu); + irq_hw_number_t hwirq; + unsigned long ipis; + unsigned int en; + + /* + * Reading enable mask does not need to be ordered as long as + * this function is called from interrupt handler because only + * the CPU itself can change it's own enable mask. + */ + en =3D atomic_read(&icpu->enable); + + /* + * Clear the IPIs we are about to handle. This pairs with the + * atomic_fetch_or_release() in ipi_mux_send_mask(). + */ + ipis =3D atomic_fetch_andnot(en, &icpu->bits) & en; + + for_each_set_bit(hwirq, &ipis, BITS_PER_TYPE(int)) + generic_handle_domain_irq(ipi_mux_domain, hwirq); +} + +/** + * ipi_mux_create - Create virtual IPIs multiplexed on top of a single + * parent IPI. + * @nr_ipi: number of virtual IPIs to create. This should + * be <=3D BITS_PER_TYPE(int) + * @mux_send: callback to trigger parent IPI for a particular CPU + * + * Returns first virq of the newly created virtual IPIs upon success + * or <=3D0 upon failure + */ +int ipi_mux_create(unsigned int nr_ipi, void (*mux_send)(unsigned int cpu)) +{ + struct fwnode_handle *fwnode; + struct irq_domain *domain; + int rc; + + if (ipi_mux_domain) + return -EEXIST; + + if (BITS_PER_TYPE(int) < nr_ipi || !mux_send) + return -EINVAL; + + ipi_mux_pcpu =3D alloc_percpu(typeof(*ipi_mux_pcpu)); + if (!ipi_mux_pcpu) + return -ENOMEM; + + fwnode =3D irq_domain_alloc_named_fwnode("IPI-Mux"); + if (!fwnode) { + pr_err("unable to create IPI Mux fwnode\n"); + rc =3D -ENOMEM; + goto fail_free_cpu; + } + + domain =3D irq_domain_create_linear(fwnode, nr_ipi, + &ipi_mux_domain_ops, NULL); + if (!domain) { + pr_err("unable to add IPI Mux domain\n"); + rc =3D -ENOMEM; + goto fail_free_fwnode; + } + + domain->flags |=3D IRQ_DOMAIN_FLAG_IPI_SINGLE; + irq_domain_update_bus_token(domain, DOMAIN_BUS_IPI); + + rc =3D __irq_domain_alloc_irqs(domain, -1, nr_ipi, + NUMA_NO_NODE, NULL, false, NULL); + if (rc <=3D 0) { + pr_err("unable to alloc IRQs from IPI Mux domain\n"); + goto fail_free_domain; + } + + ipi_mux_domain =3D domain; + ipi_mux_send =3D mux_send; + + return rc; + +fail_free_domain: + irq_domain_remove(domain); +fail_free_fwnode: + irq_domain_free_fwnode(fwnode); +fail_free_cpu: + free_percpu(ipi_mux_pcpu); + return rc; +} --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 6420BC3DA7D for ; Tue, 3 Jan 2023 14:13:16 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237855AbjACONP (ORCPT ); Tue, 3 Jan 2023 09:13:15 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59138 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237844AbjACOM5 (ORCPT ); Tue, 3 Jan 2023 09:12:57 -0500 Received: from mail-pl1-x62a.google.com (mail-pl1-x62a.google.com [IPv6:2607:f8b0:4864:20::62a]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4B0D511457 for ; Tue, 3 Jan 2023 06:12:55 -0800 (PST) Received: by mail-pl1-x62a.google.com with SMTP id d15so32705757pls.6 for ; Tue, 03 Jan 2023 06:12:55 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=nTZJdQipS216AvuSRTC92vA7PCtGAW9Kof99zlrHurk=; b=dtA5/Ba++avocp3kumNC7O+cP53rkMMtoM6BIi2bpWgcYeh6KrfBPUW3AhOUTJ2Rgp HrYQ857j7hEf2sB2tdMiZqxPCA6UFzNbAe1Ndz/tCBhkJavKW0ekFMcewTR4hFJiZgUi ydv9iHZ4re8QlzlKSTNp0BGFMcHyr1PPCVVCQV/31uw8HgRB4ti7rDAoaLdcf0EX+PdN W/ImmcCN/kYktcwQOONg/WANQQxfnWWIAebWILytVW/QsYy+DzAYu6N6NLOI1OIFPe+F jetPh2UXQw9BNGjm23ZHPsLfGL/qvgLWAkGDERhb6tK2Q49RqHl+y4ofXX3g4R9rWvM0 fN3Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=nTZJdQipS216AvuSRTC92vA7PCtGAW9Kof99zlrHurk=; b=snHyYZLgEIzNrXWXnwcIWOIlFFBaeT7PB4P7ZVULsTs9ffKq6zZNxKyiuqhv98THTj BTFTXhSsYhzQRCVC5SpciH8jimJ5lQYXoJdrZBATfP7TwTIRFgEukUeyILsfe+U0QRY+ Cm8m1hkIpO6sLK2v3XbwIj45jOzHT7x2udvUU3C9tb+3GJBj8BLwRJ27+md6xngP8Jjx 812eR3MOW7CX0NPLS3z6W0zgbQlXpqZJH+aE023e/BZ/RV/MpzZSDNcQAdoK9ssB3xr5 Lmj9X4vqeZXL9xgasgfsq0ALo8+djTAmDilDUkltPlUvXzmI/2JII7nDLOXUjCX2iHhY axMw== X-Gm-Message-State: AFqh2kpdqINH5y0BUDKPfq9GFG8CyaDgN7au/Pk+b/5Vm6qwaUicFDek 6I+iU5lr0rmcekyJ+5zeDbD7isc5OrvYFVw/ X-Google-Smtp-Source: AMrXdXvXYPfbsuCNluyperQFLyReT3aHs+SajyjDMfr29xpV5Rvae1xy0YsD6X7qlmSkLU35QmOkRg== X-Received: by 2002:a17:902:7048:b0:18f:438a:cfe1 with SMTP id h8-20020a170902704800b0018f438acfe1mr48101791plt.59.1672755174506; Tue, 03 Jan 2023 06:12:54 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.12.49 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:12:53 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel Subject: [PATCH v16 4/9] RISC-V: Treat IPIs as normal Linux IRQs Date: Tue, 3 Jan 2023 19:42:16 +0530 Message-Id: <20230103141221.772261-5-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Currently, the RISC-V kernel provides arch specific hooks (i.e. struct riscv_ipi_ops) to register IPI handling methods. The stats gathering of IPIs is also arch specific in the RISC-V kernel. Other architectures (such as ARM, ARM64, and MIPS) have moved away from custom arch specific IPI handling methods. Currently, these architectures have Linux irqchip drivers providing a range of Linux IRQ numbers to be used as IPIs and IPI triggering is done using generic IPI APIs. This approach allows architectures to treat IPIs as normal Linux IRQs and IPI stats gathering is done by the generic Linux IRQ subsystem. We extend the RISC-V IPI handling as-per above approach so that arch specific IPI handling methods (struct riscv_ipi_ops) can be removed and the IPI handling is done through the Linux IRQ subsystem. Signed-off-by: Anup Patel Acked-by: Palmer Dabbelt --- arch/riscv/Kconfig | 2 + arch/riscv/include/asm/sbi.h | 9 +- arch/riscv/include/asm/smp.h | 35 ++++--- arch/riscv/kernel/Makefile | 1 + arch/riscv/kernel/cpu-hotplug.c | 3 +- arch/riscv/kernel/irq.c | 3 +- arch/riscv/kernel/sbi-ipi.c | 77 ++++++++++++++ arch/riscv/kernel/sbi.c | 106 +++----------------- arch/riscv/kernel/smp.c | 160 +++++++++++++++--------------- arch/riscv/kernel/smpboot.c | 5 +- drivers/clocksource/timer-clint.c | 65 +++++++++--- drivers/irqchip/Kconfig | 1 + drivers/irqchip/irq-riscv-intc.c | 55 +++++----- 13 files changed, 283 insertions(+), 239 deletions(-) create mode 100644 arch/riscv/kernel/sbi-ipi.c diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig index e2b656043abf..d153e1cd890b 100644 --- a/arch/riscv/Kconfig +++ b/arch/riscv/Kconfig @@ -61,6 +61,8 @@ config RISCV select GENERIC_GETTIMEOFDAY if HAVE_GENERIC_VDSO select GENERIC_IDLE_POLL_SETUP select GENERIC_IOREMAP if MMU + select GENERIC_IRQ_IPI if SMP + select GENERIC_IRQ_IPI_MUX if SMP select GENERIC_IRQ_MULTI_HANDLER select GENERIC_IRQ_SHOW select GENERIC_IRQ_SHOW_LEVEL diff --git a/arch/riscv/include/asm/sbi.h b/arch/riscv/include/asm/sbi.h index 4ca7fbacff42..2f9db8c4aed8 100644 --- a/arch/riscv/include/asm/sbi.h +++ b/arch/riscv/include/asm/sbi.h @@ -268,8 +268,7 @@ long sbi_get_marchid(void); long sbi_get_mimpid(void); void sbi_set_timer(uint64_t stime_value); void sbi_shutdown(void); -void sbi_clear_ipi(void); -int sbi_send_ipi(const struct cpumask *cpu_mask); +void sbi_send_ipi(unsigned int cpu); int sbi_remote_fence_i(const struct cpumask *cpu_mask); int sbi_remote_sfence_vma(const struct cpumask *cpu_mask, unsigned long start, @@ -332,4 +331,10 @@ unsigned long riscv_cached_mvendorid(unsigned int cpu_= id); unsigned long riscv_cached_marchid(unsigned int cpu_id); unsigned long riscv_cached_mimpid(unsigned int cpu_id); =20 +#if IS_ENABLED(CONFIG_SMP) && IS_ENABLED(CONFIG_RISCV_SBI) +void sbi_ipi_init(void); +#else +static inline void sbi_ipi_init(void) { } +#endif + #endif /* _ASM_RISCV_SBI_H */ diff --git a/arch/riscv/include/asm/smp.h b/arch/riscv/include/asm/smp.h index 3831b638ecab..4fe7a8854c2e 100644 --- a/arch/riscv/include/asm/smp.h +++ b/arch/riscv/include/asm/smp.h @@ -15,11 +15,6 @@ struct seq_file; extern unsigned long boot_cpu_hartid; =20 -struct riscv_ipi_ops { - void (*ipi_inject)(const struct cpumask *target); - void (*ipi_clear)(void); -}; - #ifdef CONFIG_SMP /* * Mapping between linux logical cpu index and hartid. @@ -33,9 +28,6 @@ void show_ipi_stats(struct seq_file *p, int prec); /* SMP initialization hook for setup_arch */ void __init setup_smp(void); =20 -/* Called from C code, this handles an IPI. */ -void handle_IPI(struct pt_regs *regs); - /* Hook for the generic smp_call_function_many() routine. */ void arch_send_call_function_ipi_mask(struct cpumask *mask); =20 @@ -44,11 +36,17 @@ void arch_send_call_function_single_ipi(int cpu); =20 int riscv_hartid_to_cpuid(unsigned long hartid); =20 -/* Set custom IPI operations */ -void riscv_set_ipi_ops(const struct riscv_ipi_ops *ops); +/* Enable IPI for CPU hotplug */ +void riscv_ipi_enable(void); + +/* Disable IPI for CPU hotplug */ +void riscv_ipi_disable(void); =20 -/* Clear IPI for current CPU */ -void riscv_clear_ipi(void); +/* Check if IPI interrupt numbers are available */ +bool riscv_ipi_have_virq_range(void); + +/* Set the IPI interrupt numbers for arch (called by irqchip drivers) */ +void riscv_ipi_set_virq_range(int virq, int nr); =20 /* Check other CPUs stop or not */ bool smp_crash_stop_failed(void); @@ -85,11 +83,20 @@ static inline unsigned long cpuid_to_hartid_map(int cpu) return boot_cpu_hartid; } =20 -static inline void riscv_set_ipi_ops(const struct riscv_ipi_ops *ops) +static inline void riscv_ipi_enable(void) { } =20 -static inline void riscv_clear_ipi(void) +static inline void riscv_ipi_disable(void) +{ +} + +static inline bool riscv_ipi_have_virq_range(void) +{ + return false; +} + +static inline void riscv_ipi_set_virq_range(int virq, int nr) { } =20 diff --git a/arch/riscv/kernel/Makefile b/arch/riscv/kernel/Makefile index 4cf303a779ab..67f542be1bea 100644 --- a/arch/riscv/kernel/Makefile +++ b/arch/riscv/kernel/Makefile @@ -74,6 +74,7 @@ obj-$(CONFIG_PERF_EVENTS) +=3D perf_callchain.o obj-$(CONFIG_HAVE_PERF_REGS) +=3D perf_regs.o obj-$(CONFIG_RISCV_SBI) +=3D sbi.o ifeq ($(CONFIG_RISCV_SBI), y) +obj-$(CONFIG_SMP) +=3D sbi-ipi.o obj-$(CONFIG_SMP) +=3D cpu_ops_sbi.o endif obj-$(CONFIG_HOTPLUG_CPU) +=3D cpu-hotplug.o diff --git a/arch/riscv/kernel/cpu-hotplug.c b/arch/riscv/kernel/cpu-hotplu= g.c index f7a832e3a1d1..39235cf50652 100644 --- a/arch/riscv/kernel/cpu-hotplug.c +++ b/arch/riscv/kernel/cpu-hotplug.c @@ -13,7 +13,7 @@ #include #include #include -#include +#include =20 bool cpu_has_hotplug(unsigned int cpu) { @@ -43,6 +43,7 @@ int __cpu_disable(void) remove_cpu_topology(cpu); numa_remove_cpu(cpu); set_cpu_online(cpu, false); + riscv_ipi_disable(); irq_migrate_all_off_this_cpu(); =20 return ret; diff --git a/arch/riscv/kernel/irq.c b/arch/riscv/kernel/irq.c index 96d3171f0ca1..eb9a68a539e6 100644 --- a/arch/riscv/kernel/irq.c +++ b/arch/riscv/kernel/irq.c @@ -10,7 +10,7 @@ #include #include #include -#include +#include =20 static struct fwnode_handle *(*__get_intc_node)(void); =20 @@ -39,4 +39,5 @@ void __init init_IRQ(void) irqchip_init(); if (!handle_arch_irq) panic("No interrupt controller found."); + sbi_ipi_init(); } diff --git a/arch/riscv/kernel/sbi-ipi.c b/arch/riscv/kernel/sbi-ipi.c new file mode 100644 index 000000000000..41981ab88493 --- /dev/null +++ b/arch/riscv/kernel/sbi-ipi.c @@ -0,0 +1,77 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Multiplex several IPIs over a single HW IPI. + * + * Copyright (c) 2022 Ventana Micro Systems Inc. + */ + +#define pr_fmt(fmt) "riscv: " fmt +#include +#include +#include +#include +#include +#include + +static int sbi_ipi_virq; + +static void sbi_ipi_handle(struct irq_desc *desc) +{ + struct irq_chip *chip =3D irq_desc_get_chip(desc); + + chained_irq_enter(chip, desc); + + csr_clear(CSR_IP, IE_SIE); + ipi_mux_process(); + + chained_irq_exit(chip, desc); +} + +static int sbi_ipi_starting_cpu(unsigned int cpu) +{ + enable_percpu_irq(sbi_ipi_virq, irq_get_trigger_type(sbi_ipi_virq)); + return 0; +} + +void __init sbi_ipi_init(void) +{ + int virq; + struct irq_domain *domain; + + if (riscv_ipi_have_virq_range()) + return; + + domain =3D irq_find_matching_fwnode(riscv_get_intc_hwnode(), + DOMAIN_BUS_ANY); + if (!domain) { + pr_err("unable to find INTC IRQ domain\n"); + return; + } + + sbi_ipi_virq =3D irq_create_mapping(domain, RV_IRQ_SOFT); + if (!sbi_ipi_virq) { + pr_err("unable to create INTC IRQ mapping\n"); + return; + } + + virq =3D ipi_mux_create(BITS_PER_BYTE, sbi_send_ipi); + if (virq <=3D 0) { + pr_err("unable to create muxed IPIs\n"); + irq_dispose_mapping(sbi_ipi_virq); + return; + } + + irq_set_chained_handler(sbi_ipi_virq, sbi_ipi_handle); + + /* + * Don't disable IPI when CPU goes offline because + * the masking/unmasking of virtual IPIs is done + * via generic IPI-Mux + */ + cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, + "irqchip/sbi-ipi:starting", + sbi_ipi_starting_cpu, NULL); + + riscv_ipi_set_virq_range(virq, BITS_PER_BYTE); + pr_info("providing IPIs using SBI IPI extension\n"); +} diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c index ac99a70ead6a..92b9b759ab3d 100644 --- a/arch/riscv/kernel/sbi.c +++ b/arch/riscv/kernel/sbi.c @@ -17,7 +17,7 @@ unsigned long sbi_spec_version __ro_after_init =3D SBI_SP= EC_VERSION_DEFAULT; EXPORT_SYMBOL(sbi_spec_version); =20 static void (*__sbi_set_timer)(uint64_t stime) __ro_after_init; -static int (*__sbi_send_ipi)(const struct cpumask *cpu_mask) __ro_after_in= it; +static void (*__sbi_send_ipi)(unsigned int cpu) __ro_after_init; static int (*__sbi_rfence)(int fid, const struct cpumask *cpu_mask, unsigned long start, unsigned long size, unsigned long arg4, unsigned long arg5) __ro_after_init; @@ -130,17 +130,6 @@ void sbi_shutdown(void) } EXPORT_SYMBOL(sbi_shutdown); =20 -/** - * sbi_clear_ipi() - Clear any pending IPIs for the calling hart. - * - * Return: None - */ -void sbi_clear_ipi(void) -{ - sbi_ecall(SBI_EXT_0_1_CLEAR_IPI, 0, 0, 0, 0, 0, 0, 0); -} -EXPORT_SYMBOL(sbi_clear_ipi); - /** * __sbi_set_timer_v01() - Program the timer for next timer event. * @stime_value: The value after which next timer event should fire. @@ -157,17 +146,12 @@ static void __sbi_set_timer_v01(uint64_t stime_value) #endif } =20 -static int __sbi_send_ipi_v01(const struct cpumask *cpu_mask) +static void __sbi_send_ipi_v01(unsigned int cpu) { - unsigned long hart_mask; - - if (!cpu_mask || cpumask_empty(cpu_mask)) - cpu_mask =3D cpu_online_mask; - hart_mask =3D __sbi_v01_cpumask_to_hartmask(cpu_mask); - + unsigned long hart_mask =3D + __sbi_v01_cpumask_to_hartmask(cpumask_of(cpu)); sbi_ecall(SBI_EXT_0_1_SEND_IPI, 0, (unsigned long)(&hart_mask), 0, 0, 0, 0, 0); - return 0; } =20 static int __sbi_rfence_v01(int fid, const struct cpumask *cpu_mask, @@ -216,12 +200,10 @@ static void __sbi_set_timer_v01(uint64_t stime_value) sbi_major_version(), sbi_minor_version()); } =20 -static int __sbi_send_ipi_v01(const struct cpumask *cpu_mask) +static void __sbi_send_ipi_v01(unsigned int cpu) { pr_warn("IPI extension is not available in SBI v%lu.%lu\n", sbi_major_version(), sbi_minor_version()); - - return 0; } =20 static int __sbi_rfence_v01(int fid, const struct cpumask *cpu_mask, @@ -248,55 +230,18 @@ static void __sbi_set_timer_v02(uint64_t stime_value) #endif } =20 -static int __sbi_send_ipi_v02(const struct cpumask *cpu_mask) +static void __sbi_send_ipi_v02(unsigned int cpu) { - unsigned long hartid, cpuid, hmask =3D 0, hbase =3D 0, htop =3D 0; - struct sbiret ret =3D {0}; int result; + struct sbiret ret =3D {0}; =20 - if (!cpu_mask || cpumask_empty(cpu_mask)) - cpu_mask =3D cpu_online_mask; - - for_each_cpu(cpuid, cpu_mask) { - hartid =3D cpuid_to_hartid_map(cpuid); - if (hmask) { - if (hartid + BITS_PER_LONG <=3D htop || - hbase + BITS_PER_LONG <=3D hartid) { - ret =3D sbi_ecall(SBI_EXT_IPI, - SBI_EXT_IPI_SEND_IPI, hmask, - hbase, 0, 0, 0, 0); - if (ret.error) - goto ecall_failed; - hmask =3D 0; - } else if (hartid < hbase) { - /* shift the mask to fit lower hartid */ - hmask <<=3D hbase - hartid; - hbase =3D hartid; - } - } - if (!hmask) { - hbase =3D hartid; - htop =3D hartid; - } else if (hartid > htop) { - htop =3D hartid; - } - hmask |=3D BIT(hartid - hbase); - } - - if (hmask) { - ret =3D sbi_ecall(SBI_EXT_IPI, SBI_EXT_IPI_SEND_IPI, - hmask, hbase, 0, 0, 0, 0); - if (ret.error) - goto ecall_failed; + ret =3D sbi_ecall(SBI_EXT_IPI, SBI_EXT_IPI_SEND_IPI, + 1UL, cpuid_to_hartid_map(cpu), 0, 0, 0, 0); + if (ret.error) { + result =3D sbi_err_map_linux_errno(ret.error); + pr_err("%s: hbase =3D [%lu] failed (error [%d])\n", + __func__, cpuid_to_hartid_map(cpu), result); } - - return 0; - -ecall_failed: - result =3D sbi_err_map_linux_errno(ret.error); - pr_err("%s: hbase =3D [%lu] hmask =3D [0x%lx] failed (error [%d])\n", - __func__, hbase, hmask, result); - return result; } =20 static int __sbi_rfence_v02_call(unsigned long fid, unsigned long hmask, @@ -410,13 +355,11 @@ void sbi_set_timer(uint64_t stime_value) =20 /** * sbi_send_ipi() - Send an IPI to any hart. - * @cpu_mask: A cpu mask containing all the target harts. - * - * Return: 0 on success, appropriate linux error code otherwise. + * @cpu: Logical id of the target CPU. */ -int sbi_send_ipi(const struct cpumask *cpu_mask) +void sbi_send_ipi(unsigned int cpu) { - return __sbi_send_ipi(cpu_mask); + __sbi_send_ipi(cpu); } EXPORT_SYMBOL(sbi_send_ipi); =20 @@ -641,21 +584,6 @@ long sbi_get_mimpid(void) } EXPORT_SYMBOL_GPL(sbi_get_mimpid); =20 -static void sbi_send_cpumask_ipi(const struct cpumask *target) -{ - sbi_send_ipi(target); -} - -static void sbi_ipi_clear(void) -{ - csr_clear(CSR_IP, IE_SIE); -} - -static const struct riscv_ipi_ops sbi_ipi_ops =3D { - .ipi_inject =3D sbi_send_cpumask_ipi, - .ipi_clear =3D sbi_ipi_clear -}; - void __init sbi_init(void) { int ret; @@ -702,6 +630,4 @@ void __init sbi_init(void) __sbi_send_ipi =3D __sbi_send_ipi_v01; __sbi_rfence =3D __sbi_rfence_v01; } - - riscv_set_ipi_ops(&sbi_ipi_ops); } diff --git a/arch/riscv/kernel/smp.c b/arch/riscv/kernel/smp.c index 8a12768c09ee..47e7ecfedb4d 100644 --- a/arch/riscv/kernel/smp.c +++ b/arch/riscv/kernel/smp.c @@ -13,14 +13,15 @@ #include #include #include +#include #include #include #include #include #include +#include #include =20 -#include #include #include #include @@ -44,11 +45,10 @@ void __init smp_setup_processor_id(void) cpuid_to_hartid_map(0) =3D boot_cpu_hartid; } =20 -/* A collection of single bit ipi messages. */ -static struct { - unsigned long stats[IPI_MAX] ____cacheline_aligned; - unsigned long bits ____cacheline_aligned; -} ipi_data[NR_CPUS] __cacheline_aligned; +static DEFINE_PER_CPU_READ_MOSTLY(int, ipi_dummy_dev); +static int ipi_virq_base __ro_after_init; +static int nr_ipi __ro_after_init =3D IPI_MAX; +static struct irq_desc *ipi_desc[IPI_MAX] __read_mostly; =20 int riscv_hartid_to_cpuid(unsigned long hartid) { @@ -100,46 +100,14 @@ static inline void ipi_cpu_crash_stop(unsigned int cp= u, struct pt_regs *regs) } #endif =20 -static const struct riscv_ipi_ops *ipi_ops __ro_after_init; - -void riscv_set_ipi_ops(const struct riscv_ipi_ops *ops) -{ - ipi_ops =3D ops; -} -EXPORT_SYMBOL_GPL(riscv_set_ipi_ops); - -void riscv_clear_ipi(void) -{ - if (ipi_ops && ipi_ops->ipi_clear) - ipi_ops->ipi_clear(); -} -EXPORT_SYMBOL_GPL(riscv_clear_ipi); - static void send_ipi_mask(const struct cpumask *mask, enum ipi_message_typ= e op) { - int cpu; - - smp_mb__before_atomic(); - for_each_cpu(cpu, mask) - set_bit(op, &ipi_data[cpu].bits); - smp_mb__after_atomic(); - - if (ipi_ops && ipi_ops->ipi_inject) - ipi_ops->ipi_inject(mask); - else - pr_warn("SMP: IPI inject method not available\n"); + __ipi_send_mask(ipi_desc[op], mask); } =20 static void send_ipi_single(int cpu, enum ipi_message_type op) { - smp_mb__before_atomic(); - set_bit(op, &ipi_data[cpu].bits); - smp_mb__after_atomic(); - - if (ipi_ops && ipi_ops->ipi_inject) - ipi_ops->ipi_inject(cpumask_of(cpu)); - else - pr_warn("SMP: IPI inject method not available\n"); + __ipi_send_mask(ipi_desc[op], cpumask_of(cpu)); } =20 #ifdef CONFIG_IRQ_WORK @@ -149,59 +117,89 @@ void arch_irq_work_raise(void) } #endif =20 -void handle_IPI(struct pt_regs *regs) +static irqreturn_t handle_IPI(int irq, void *data) +{ + int ipi =3D irq - ipi_virq_base; + + switch (ipi) { + case IPI_RESCHEDULE: + scheduler_ipi(); + break; + case IPI_CALL_FUNC: + generic_smp_call_function_interrupt(); + break; + case IPI_CPU_STOP: + ipi_stop(); + break; + case IPI_CPU_CRASH_STOP: + ipi_cpu_crash_stop(smp_processor_id(), get_irq_regs()); + break; + case IPI_IRQ_WORK: + irq_work_run(); + break; +#ifdef CONFIG_GENERIC_CLOCKEVENTS_BROADCAST + case IPI_TIMER: + tick_receive_broadcast(); + break; +#endif + default: + pr_warn("CPU%d: unhandled IPI%d\n", smp_processor_id(), ipi); + break; + } + + return IRQ_HANDLED; +} + +void riscv_ipi_enable(void) { - unsigned int cpu =3D smp_processor_id(); - unsigned long *pending_ipis =3D &ipi_data[cpu].bits; - unsigned long *stats =3D ipi_data[cpu].stats; + int i; =20 - riscv_clear_ipi(); + if (WARN_ON_ONCE(!ipi_virq_base)) + return; =20 - while (true) { - unsigned long ops; + for (i =3D 0; i < nr_ipi; i++) + enable_percpu_irq(ipi_virq_base + i, 0); +} =20 - /* Order bit clearing and data access. */ - mb(); +void riscv_ipi_disable(void) +{ + int i; =20 - ops =3D xchg(pending_ipis, 0); - if (ops =3D=3D 0) - return; + if (WARN_ON_ONCE(!ipi_virq_base)) + return; =20 - if (ops & (1 << IPI_RESCHEDULE)) { - stats[IPI_RESCHEDULE]++; - scheduler_ipi(); - } + for (i =3D 0; i < nr_ipi; i++) + disable_percpu_irq(ipi_virq_base + i); +} =20 - if (ops & (1 << IPI_CALL_FUNC)) { - stats[IPI_CALL_FUNC]++; - generic_smp_call_function_interrupt(); - } +bool riscv_ipi_have_virq_range(void) +{ + return (ipi_virq_base) ? true : false; +} =20 - if (ops & (1 << IPI_CPU_STOP)) { - stats[IPI_CPU_STOP]++; - ipi_stop(); - } +void riscv_ipi_set_virq_range(int virq, int nr) +{ + int i, err; =20 - if (ops & (1 << IPI_CPU_CRASH_STOP)) { - ipi_cpu_crash_stop(cpu, get_irq_regs()); - } + if (WARN_ON(ipi_virq_base)) + return; =20 - if (ops & (1 << IPI_IRQ_WORK)) { - stats[IPI_IRQ_WORK]++; - irq_work_run(); - } + WARN_ON(nr < IPI_MAX); + nr_ipi =3D min(nr, IPI_MAX); + ipi_virq_base =3D virq; =20 -#ifdef CONFIG_GENERIC_CLOCKEVENTS_BROADCAST - if (ops & (1 << IPI_TIMER)) { - stats[IPI_TIMER]++; - tick_receive_broadcast(); - } -#endif - BUG_ON((ops >> IPI_MAX) !=3D 0); + /* Request IPIs */ + for (i =3D 0; i < nr_ipi; i++) { + err =3D request_percpu_irq(ipi_virq_base + i, handle_IPI, + "IPI", &ipi_dummy_dev); + WARN_ON(err); =20 - /* Order data access and bit testing. */ - mb(); + ipi_desc[i] =3D irq_to_desc(ipi_virq_base + i); + irq_set_status_flags(ipi_virq_base + i, IRQ_HIDDEN); } + + /* Enabled IPIs for boot CPU immediately */ + riscv_ipi_enable(); } =20 static const char * const ipi_names[] =3D { @@ -221,7 +219,7 @@ void show_ipi_stats(struct seq_file *p, int prec) seq_printf(p, "%*s%u:%s", prec - 1, "IPI", i, prec >=3D 4 ? " " : ""); for_each_online_cpu(cpu) - seq_printf(p, "%10lu ", ipi_data[cpu].stats[i]); + seq_printf(p, "%10u ", irq_desc_kstat_cpu(ipi_desc[i], cpu)); seq_printf(p, " %s\n", ipi_names[i]); } } diff --git a/arch/riscv/kernel/smpboot.c b/arch/riscv/kernel/smpboot.c index 3373df413c88..9cbdb960515b 100644 --- a/arch/riscv/kernel/smpboot.c +++ b/arch/riscv/kernel/smpboot.c @@ -30,7 +30,6 @@ #include #include #include -#include #include =20 #include "head.h" @@ -157,12 +156,12 @@ asmlinkage __visible void smp_callin(void) struct mm_struct *mm =3D &init_mm; unsigned int curr_cpuid =3D smp_processor_id(); =20 - riscv_clear_ipi(); - /* All kernel threads share the same mm context. */ mmgrab(mm); current->active_mm =3D mm; =20 + riscv_ipi_enable(); + store_cpu_topology(curr_cpuid); notify_cpu_starting(curr_cpuid); numa_add_cpu(curr_cpuid); diff --git a/drivers/clocksource/timer-clint.c b/drivers/clocksource/timer-= clint.c index 6cfe2ab73eb0..7ccc16dd6a76 100644 --- a/drivers/clocksource/timer-clint.c +++ b/drivers/clocksource/timer-clint.c @@ -17,6 +17,9 @@ #include #include #include +#include +#include +#include #include #include #include @@ -31,6 +34,7 @@ =20 /* CLINT manages IPI and Timer for RISC-V M-mode */ static u32 __iomem *clint_ipi_base; +static unsigned int clint_ipi_irq; static u64 __iomem *clint_timer_cmp; static u64 __iomem *clint_timer_val; static unsigned long clint_timer_freq; @@ -41,12 +45,10 @@ u64 __iomem *clint_time_val; EXPORT_SYMBOL(clint_time_val); #endif =20 -static void clint_send_ipi(const struct cpumask *target) +#ifdef CONFIG_SMP +static void clint_send_ipi(unsigned int cpu) { - unsigned int cpu; - - for_each_cpu(cpu, target) - writel(1, clint_ipi_base + cpuid_to_hartid_map(cpu)); + writel(1, clint_ipi_base + cpuid_to_hartid_map(cpu)); } =20 static void clint_clear_ipi(void) @@ -54,10 +56,18 @@ static void clint_clear_ipi(void) writel(0, clint_ipi_base + cpuid_to_hartid_map(smp_processor_id())); } =20 -static struct riscv_ipi_ops clint_ipi_ops =3D { - .ipi_inject =3D clint_send_ipi, - .ipi_clear =3D clint_clear_ipi, -}; +static void clint_ipi_interrupt(struct irq_desc *desc) +{ + struct irq_chip *chip =3D irq_desc_get_chip(desc); + + chained_irq_enter(chip, desc); + + clint_clear_ipi(); + ipi_mux_process(); + + chained_irq_exit(chip, desc); +} +#endif =20 #ifdef CONFIG_64BIT #define clint_get_cycles() readq_relaxed(clint_timer_val) @@ -125,12 +135,19 @@ static int clint_timer_starting_cpu(unsigned int cpu) =20 enable_percpu_irq(clint_timer_irq, irq_get_trigger_type(clint_timer_irq)); + enable_percpu_irq(clint_ipi_irq, + irq_get_trigger_type(clint_ipi_irq)); return 0; } =20 static int clint_timer_dying_cpu(unsigned int cpu) { disable_percpu_irq(clint_timer_irq); + /* + * Don't disable IPI when CPU goes offline because + * the masking/unmasking of virtual IPIs is done + * via generic IPI-Mux + */ return 0; } =20 @@ -170,6 +187,12 @@ static int __init clint_timer_init_dt(struct device_no= de *np) return -ENODEV; } =20 + /* Find parent irq domain and map ipi irq */ + if (!clint_ipi_irq && + oirq.args[0] =3D=3D RV_IRQ_SOFT && + irq_find_host(oirq.np)) + clint_ipi_irq =3D irq_of_parse_and_map(np, i); + /* Find parent irq domain and map timer irq */ if (!clint_timer_irq && oirq.args[0] =3D=3D RV_IRQ_TIMER && @@ -177,9 +200,9 @@ static int __init clint_timer_init_dt(struct device_nod= e *np) clint_timer_irq =3D irq_of_parse_and_map(np, i); } =20 - /* If CLINT timer irq not found then fail */ - if (!clint_timer_irq) { - pr_err("%pOFP: timer irq not found\n", np); + /* If CLINT ipi or timer irq not found then fail */ + if (!clint_ipi_irq || !clint_timer_irq) { + pr_err("%pOFP: ipi/timer irq not found\n", np); return -ENODEV; } =20 @@ -219,6 +242,19 @@ static int __init clint_timer_init_dt(struct device_no= de *np) goto fail_iounmap; } =20 +#ifdef CONFIG_SMP + rc =3D ipi_mux_create(BITS_PER_BYTE, clint_send_ipi); + if (rc <=3D 0) { + pr_err("unable to create muxed IPIs\n"); + rc =3D (rc < 0) ? rc : -ENODEV; + goto fail_free_irq; + } + + irq_set_chained_handler(clint_ipi_irq, clint_ipi_interrupt); + riscv_ipi_set_virq_range(rc, BITS_PER_BYTE); + clint_clear_ipi(); +#endif + rc =3D cpuhp_setup_state(CPUHP_AP_CLINT_TIMER_STARTING, "clockevents/clint/timer:starting", clint_timer_starting_cpu, @@ -228,13 +264,10 @@ static int __init clint_timer_init_dt(struct device_n= ode *np) goto fail_free_irq; } =20 - riscv_set_ipi_ops(&clint_ipi_ops); - clint_clear_ipi(); - return 0; =20 fail_free_irq: - free_irq(clint_timer_irq, &clint_clock_event); + free_percpu_irq(clint_timer_irq, &clint_clock_event); fail_iounmap: iounmap(base); return rc; diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig index caa952c40ff9..0e3a7749d87c 100644 --- a/drivers/irqchip/Kconfig +++ b/drivers/irqchip/Kconfig @@ -540,6 +540,7 @@ config TI_PRUSS_INTC config RISCV_INTC bool depends on RISCV + select IRQ_DOMAIN_HIERARCHY =20 config SIFIVE_PLIC bool diff --git a/drivers/irqchip/irq-riscv-intc.c b/drivers/irqchip/irq-riscv-i= ntc.c index 9066467e99e4..784d25645704 100644 --- a/drivers/irqchip/irq-riscv-intc.c +++ b/drivers/irqchip/irq-riscv-intc.c @@ -26,20 +26,7 @@ static asmlinkage void riscv_intc_irq(struct pt_regs *re= gs) if (unlikely(cause >=3D BITS_PER_LONG)) panic("unexpected interrupt cause"); =20 - switch (cause) { -#ifdef CONFIG_SMP - case RV_IRQ_SOFT: - /* - * We only use software interrupts to pass IPIs, so if a - * non-SMP system gets one, then we don't know what to do. - */ - handle_IPI(regs); - break; -#endif - default: - generic_handle_domain_irq(intc_domain, cause); - break; - } + generic_handle_domain_irq(intc_domain, cause); } =20 /* @@ -59,18 +46,6 @@ static void riscv_intc_irq_unmask(struct irq_data *d) csr_set(CSR_IE, BIT(d->hwirq)); } =20 -static int riscv_intc_cpu_starting(unsigned int cpu) -{ - csr_set(CSR_IE, BIT(RV_IRQ_SOFT)); - return 0; -} - -static int riscv_intc_cpu_dying(unsigned int cpu) -{ - csr_clear(CSR_IE, BIT(RV_IRQ_SOFT)); - return 0; -} - static struct irq_chip riscv_intc_chip =3D { .name =3D "RISC-V INTC", .irq_mask =3D riscv_intc_irq_mask, @@ -87,9 +62,32 @@ static int riscv_intc_domain_map(struct irq_domain *d, u= nsigned int irq, return 0; } =20 +static int riscv_intc_domain_alloc(struct irq_domain *domain, + unsigned int virq, unsigned int nr_irqs, + void *arg) +{ + int i, ret; + irq_hw_number_t hwirq; + unsigned int type =3D IRQ_TYPE_NONE; + struct irq_fwspec *fwspec =3D arg; + + ret =3D irq_domain_translate_onecell(domain, fwspec, &hwirq, &type); + if (ret) + return ret; + + for (i =3D 0; i < nr_irqs; i++) { + ret =3D riscv_intc_domain_map(domain, virq + i, hwirq + i); + if (ret) + return ret; + } + + return 0; +} + static const struct irq_domain_ops riscv_intc_domain_ops =3D { .map =3D riscv_intc_domain_map, .xlate =3D irq_domain_xlate_onecell, + .alloc =3D riscv_intc_domain_alloc }; =20 static struct fwnode_handle *riscv_intc_hwnode(void) @@ -133,11 +131,6 @@ static int __init riscv_intc_init(struct device_node *= node, =20 riscv_set_intc_hwnode_fn(riscv_intc_hwnode); =20 - cpuhp_setup_state(CPUHP_AP_IRQ_RISCV_STARTING, - "irqchip/riscv/intc:starting", - riscv_intc_cpu_starting, - riscv_intc_cpu_dying); - pr_info("%d local interrupts mapped\n", BITS_PER_LONG); =20 return 0; --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 14351C3DA7D for ; Tue, 3 Jan 2023 14:13:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237860AbjACONR (ORCPT ); Tue, 3 Jan 2023 09:13:17 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58696 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237847AbjACONC (ORCPT ); Tue, 3 Jan 2023 09:13:02 -0500 Received: from mail-pl1-x62e.google.com (mail-pl1-x62e.google.com [IPv6:2607:f8b0:4864:20::62e]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id CDE4411470 for ; Tue, 3 Jan 2023 06:13:00 -0800 (PST) Received: by mail-pl1-x62e.google.com with SMTP id jl4so26426473plb.8 for ; Tue, 03 Jan 2023 06:13:00 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=LpOGLHCCze3Qrz0T2fIbHYDVqeXz60uKcyL+jWZhlg8=; b=UYZ4t90Kwc4TJhHwVS4LeCURIv27178l+3wPhgqNJrChuixgNRsyoV2x7DAQqtF57r X9XApiI4KUt+GwjAdvs6HJVo5NHQ4vRg55KQgyCOdtIsLrnT9sgeJuYQhR/2XweG69Z2 rMjbyT/DjuVFkgYljXkWSeTBTqaQsriid+eqgBnAg/LA6g3+DIQALDa8JMikZadf3WJg g4eKw0fqPWMq5dZkRabV46Zii8SMq1K6G29d4SHfjwYqtXGp9tvxKcE5CsJGr8MWhHxn RK/n+7srKVAoYZqsLiorHeFQUPeMwR3oiINN8n/wQotWQ+oHPgh1yLGP7P67oO36RaqA wQ1Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=LpOGLHCCze3Qrz0T2fIbHYDVqeXz60uKcyL+jWZhlg8=; b=DtYnEM6cWMishaKB4+y1JyHZd4NRokI4Rfl3hem36o9v8n0jH7hEWvG3MvLs+vzoix WQKtQQ7lfkImFfDEyOhA9xnMMwzp4Af3wU3cdP0WBA0DCIVc80y5Y2nJX1ZvmuXOd0Qu UM4rE0vtGnAObTid8FrUGP+e4NJCw9uSquPgpqASU2MuSahUPkZK4iz2ZormPQ57tGxM OShHcrzodsA3wuFzQmtVpNrU9w1gRAMrc5fE9D2R5wltayy5tG6mbbJxqwTrOMX7asXX 3JkW8lIhTJ7aDAbANO7E/AiJj3MYIcbIkAxaes2DQ5PusFrBr/GkRXZ45LBB+Gnt57Yk Gj9Q== X-Gm-Message-State: AFqh2krji5SxOvyhbBGvU4Mly1dIV2US10roxiP9JNquVCA5G/pYonvN WA/doYLbPKwNab5x3dw6jel9Wg== X-Google-Smtp-Source: AMrXdXvBmVydsjqVigHSLQaUkXoDfPacX7mbai5MozoCC4s8JLE/D1PuAYNnEchkfa6UqfNL+smqbg== X-Received: by 2002:a17:902:654e:b0:191:14c1:626c with SMTP id d14-20020a170902654e00b0019114c1626cmr50486044pln.9.1672755180248; Tue, 03 Jan 2023 06:13:00 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.12.54 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:12:59 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel , Atish Patra Subject: [PATCH v16 5/9] RISC-V: Allow marking IPIs as suitable for remote FENCEs Date: Tue, 3 Jan 2023 19:42:17 +0530 Message-Id: <20230103141221.772261-6-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" To do remote FENCEs (i.e. remote TLB flushes) using IPI calls on the RISC-V kernel, we need hardware mechanism to directly inject IPI from the supervisor mode (i.e. RISC-V kernel) instead of using SBI calls. The upcoming AIA IMSIC devices allow direct IPI injection from the supervisor mode (i.e. RISC-V kernel). To support this, we extend the riscv_ipi_set_virq_range() function so that IPI provider (i.e. irqchip drivers can mark IPIs as suitable for remote FENCEs. Signed-off-by: Anup Patel Reviewed-by: Atish Patra Acked-by: Palmer Dabbelt --- arch/riscv/include/asm/smp.h | 18 ++++++++++++++++-- arch/riscv/kernel/sbi-ipi.c | 2 +- arch/riscv/kernel/smp.c | 11 ++++++++++- drivers/clocksource/timer-clint.c | 2 +- 4 files changed, 28 insertions(+), 5 deletions(-) diff --git a/arch/riscv/include/asm/smp.h b/arch/riscv/include/asm/smp.h index 4fe7a8854c2e..c4b77017ec58 100644 --- a/arch/riscv/include/asm/smp.h +++ b/arch/riscv/include/asm/smp.h @@ -16,6 +16,9 @@ struct seq_file; extern unsigned long boot_cpu_hartid; =20 #ifdef CONFIG_SMP + +#include + /* * Mapping between linux logical cpu index and hartid. */ @@ -46,7 +49,12 @@ void riscv_ipi_disable(void); bool riscv_ipi_have_virq_range(void); =20 /* Set the IPI interrupt numbers for arch (called by irqchip drivers) */ -void riscv_ipi_set_virq_range(int virq, int nr); +void riscv_ipi_set_virq_range(int virq, int nr, bool use_for_rfence); + +/* Check if we can use IPIs for remote FENCEs */ +DECLARE_STATIC_KEY_FALSE(riscv_ipi_for_rfence); +#define riscv_use_ipi_for_rfence() \ + static_branch_unlikely(&riscv_ipi_for_rfence) =20 /* Check other CPUs stop or not */ bool smp_crash_stop_failed(void); @@ -96,10 +104,16 @@ static inline bool riscv_ipi_have_virq_range(void) return false; } =20 -static inline void riscv_ipi_set_virq_range(int virq, int nr) +static inline void riscv_ipi_set_virq_range(int virq, int nr, + bool use_for_rfence) { } =20 +static inline bool riscv_use_ipi_for_rfence(void) +{ + return false; +} + #endif /* CONFIG_SMP */ =20 #if defined(CONFIG_HOTPLUG_CPU) && (CONFIG_SMP) diff --git a/arch/riscv/kernel/sbi-ipi.c b/arch/riscv/kernel/sbi-ipi.c index 41981ab88493..a4559695ce62 100644 --- a/arch/riscv/kernel/sbi-ipi.c +++ b/arch/riscv/kernel/sbi-ipi.c @@ -72,6 +72,6 @@ void __init sbi_ipi_init(void) "irqchip/sbi-ipi:starting", sbi_ipi_starting_cpu, NULL); =20 - riscv_ipi_set_virq_range(virq, BITS_PER_BYTE); + riscv_ipi_set_virq_range(virq, BITS_PER_BYTE, false); pr_info("providing IPIs using SBI IPI extension\n"); } diff --git a/arch/riscv/kernel/smp.c b/arch/riscv/kernel/smp.c index 47e7ecfedb4d..5f985a197eff 100644 --- a/arch/riscv/kernel/smp.c +++ b/arch/riscv/kernel/smp.c @@ -177,7 +177,10 @@ bool riscv_ipi_have_virq_range(void) return (ipi_virq_base) ? true : false; } =20 -void riscv_ipi_set_virq_range(int virq, int nr) +DEFINE_STATIC_KEY_FALSE(riscv_ipi_for_rfence); +EXPORT_SYMBOL_GPL(riscv_ipi_for_rfence); + +void riscv_ipi_set_virq_range(int virq, int nr, bool use_for_rfence) { int i, err; =20 @@ -200,6 +203,12 @@ void riscv_ipi_set_virq_range(int virq, int nr) =20 /* Enabled IPIs for boot CPU immediately */ riscv_ipi_enable(); + + /* Update RFENCE static key */ + if (use_for_rfence) + static_branch_enable(&riscv_ipi_for_rfence); + else + static_branch_disable(&riscv_ipi_for_rfence); } =20 static const char * const ipi_names[] =3D { diff --git a/drivers/clocksource/timer-clint.c b/drivers/clocksource/timer-= clint.c index 7ccc16dd6a76..9a55e733ae99 100644 --- a/drivers/clocksource/timer-clint.c +++ b/drivers/clocksource/timer-clint.c @@ -251,7 +251,7 @@ static int __init clint_timer_init_dt(struct device_nod= e *np) } =20 irq_set_chained_handler(clint_ipi_irq, clint_ipi_interrupt); - riscv_ipi_set_virq_range(rc, BITS_PER_BYTE); + riscv_ipi_set_virq_range(rc, BITS_PER_BYTE, true); clint_clear_ipi(); #endif =20 --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 83D43C3DA7D for ; Tue, 3 Jan 2023 14:13:23 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237865AbjACONW (ORCPT ); Tue, 3 Jan 2023 09:13:22 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59548 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237815AbjACONH (ORCPT ); Tue, 3 Jan 2023 09:13:07 -0500 Received: from mail-pj1-x102c.google.com (mail-pj1-x102c.google.com [IPv6:2607:f8b0:4864:20::102c]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 25A3F1117B for ; Tue, 3 Jan 2023 06:13:06 -0800 (PST) Received: by mail-pj1-x102c.google.com with SMTP id o31-20020a17090a0a2200b00223fedffb30so31209104pjo.3 for ; Tue, 03 Jan 2023 06:13:06 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=UyYgBnGIgCz0adPXlD9wrM9sYGYzSlN2q//39kO7eCQ=; b=VFrCfQGvUaFBTByIk9Fw+SoRN2adKAy5nq1pk3GknMGICwVIeq6PqVcguH2UT7fw14 scTfEiPCmS5EzqItFCCz3rcibVJcuNQ2y5jWui8Whhs1MAG7NkznD1dP+mLOxBglEv/Q gmleGq2bTYjk+2DCP2SnIOYTMUk/rkE5BTphu1CXzN7nSQxD6hc2f9QyvgaqcHYR1t/q zAMczPc1mVjoTPuXRUXL90hUwUJJ0GvIiejgH+mIbnfVvlDT7OivXJm00f8S9D+AhWF0 3+PLUFoH+qxDSD5zwV7z3ThGwTPIeCrpIN/mI6txHvOc0J7WxwvvKSwQ/G2/6uzfapKl 5jkg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=UyYgBnGIgCz0adPXlD9wrM9sYGYzSlN2q//39kO7eCQ=; b=bGVfLIqLsEdCAC7+pigTk7/lABjYORGDmaXTY84Lvc1mnWyV1OwONtHX+XTSkn+Dxl hgvLV9FYmiemIS2a7mMye6rNhuU556lsBYws4dMXMtTyrF7tWiUidEgjLfq8LzilY0JV anl6SHqp1ixMBXW3wn03gWkcPHCwBjpQa7mukMwti9p+doAof4nMpr79sjuv1F8M+O80 HEJbc1qOn8kXjzVIFfjmpes14RJn9ZD6cmRs4MOdWHCnB5fOAqMsCTpNZ0wcroz8gK97 UCXUTovWJPEx3eW43s4VQl7Rn/OyDJ2no5jmAhd4U2qateII5yh6+Mlk5wR+xCYVbCNX ywlA== X-Gm-Message-State: AFqh2krgd+pb6i5mS8HFIJudjlHU9HRPYgekSBheKV7NimQFOWBaATMO dYl+IAfuRiO2SgSdwJIi0tf4Ng== X-Google-Smtp-Source: AMrXdXttb+oVh5eibKiteLalBjeH5MUTIPsfMvziRRV5ZsaGNN266xnWVWLFBygP/lCOLay51PTgqg== X-Received: by 2002:a17:903:200b:b0:192:467e:7379 with SMTP id s11-20020a170903200b00b00192467e7379mr39572832pla.49.1672755185537; Tue, 03 Jan 2023 06:13:05 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.13.00 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:13:05 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel , Atish Patra Subject: [PATCH v16 6/9] RISC-V: Use IPIs for remote TLB flush when possible Date: Tue, 3 Jan 2023 19:42:18 +0530 Message-Id: <20230103141221.772261-7-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" If we have specialized interrupt controller (such as AIA IMSIC) which allows supervisor mode to directly inject IPIs without any assistance from M-mode or HS-mode then using such specialized interrupt controller, we can do remote TLB flushes directly from supervisor mode instead of using the SBI RFENCE calls. This patch extends remote TLB flush functions to use supervisor mode IPIs whenever direct supervisor mode IPIs.are supported by interrupt controller. Signed-off-by: Anup Patel Reviewed-by: Atish Patra Acked-by: Palmer Dabbelt --- arch/riscv/mm/tlbflush.c | 93 +++++++++++++++++++++++++++++++++------- 1 file changed, 78 insertions(+), 15 deletions(-) diff --git a/arch/riscv/mm/tlbflush.c b/arch/riscv/mm/tlbflush.c index ce7dfc81bb3f..91078377344d 100644 --- a/arch/riscv/mm/tlbflush.c +++ b/arch/riscv/mm/tlbflush.c @@ -7,14 +7,62 @@ #include #include =20 +static inline void local_flush_tlb_range(unsigned long start, + unsigned long size, unsigned long stride) +{ + if (size <=3D stride) + local_flush_tlb_page(start); + else + local_flush_tlb_all(); +} + +static inline void local_flush_tlb_range_asid(unsigned long start, + unsigned long size, unsigned long stride, unsigned long asid) +{ + if (size <=3D stride) + local_flush_tlb_page_asid(start, asid); + else + local_flush_tlb_all_asid(asid); +} + +static void __ipi_flush_tlb_all(void *info) +{ + local_flush_tlb_all(); +} + void flush_tlb_all(void) { - sbi_remote_sfence_vma(NULL, 0, -1); + if (riscv_use_ipi_for_rfence()) + on_each_cpu(__ipi_flush_tlb_all, NULL, 1); + else + sbi_remote_sfence_vma(NULL, 0, -1); +} + +struct flush_tlb_range_data { + unsigned long asid; + unsigned long start; + unsigned long size; + unsigned long stride; +}; + +static void __ipi_flush_tlb_range_asid(void *info) +{ + struct flush_tlb_range_data *d =3D info; + + local_flush_tlb_range_asid(d->start, d->size, d->stride, d->asid); +} + +static void __ipi_flush_tlb_range(void *info) +{ + struct flush_tlb_range_data *d =3D info; + + local_flush_tlb_range(d->start, d->size, d->stride); } =20 -static void __sbi_tlb_flush_range(struct mm_struct *mm, unsigned long star= t, - unsigned long size, unsigned long stride) +static void __flush_tlb_range(struct mm_struct *mm, unsigned long start, + unsigned long size, unsigned long stride) { + struct flush_tlb_range_data ftd; struct cpumask *pmask =3D &mm->context.tlb_stale_mask; struct cpumask *cmask =3D mm_cpumask(mm); unsigned int cpuid; @@ -39,19 +87,34 @@ static void __sbi_tlb_flush_range(struct mm_struct *mm,= unsigned long start, cpumask_andnot(pmask, pmask, cmask); =20 if (broadcast) { - sbi_remote_sfence_vma_asid(cmask, start, size, asid); - } else if (size <=3D stride) { - local_flush_tlb_page_asid(start, asid); + if (riscv_use_ipi_for_rfence()) { + ftd.asid =3D asid; + ftd.start =3D start; + ftd.size =3D size; + ftd.stride =3D stride; + on_each_cpu_mask(cmask, + __ipi_flush_tlb_range_asid, + &ftd, 1); + } else + sbi_remote_sfence_vma_asid(cmask, + start, size, asid); } else { - local_flush_tlb_all_asid(asid); + local_flush_tlb_range_asid(start, size, stride, asid); } } else { if (broadcast) { - sbi_remote_sfence_vma(cmask, start, size); - } else if (size <=3D stride) { - local_flush_tlb_page(start); + if (riscv_use_ipi_for_rfence()) { + ftd.asid =3D 0; + ftd.start =3D start; + ftd.size =3D size; + ftd.stride =3D stride; + on_each_cpu_mask(cmask, + __ipi_flush_tlb_range, + &ftd, 1); + } else + sbi_remote_sfence_vma(cmask, start, size); } else { - local_flush_tlb_all(); + local_flush_tlb_range(start, size, stride); } } =20 @@ -60,23 +123,23 @@ static void __sbi_tlb_flush_range(struct mm_struct *mm= , unsigned long start, =20 void flush_tlb_mm(struct mm_struct *mm) { - __sbi_tlb_flush_range(mm, 0, -1, PAGE_SIZE); + __flush_tlb_range(mm, 0, -1, PAGE_SIZE); } =20 void flush_tlb_page(struct vm_area_struct *vma, unsigned long addr) { - __sbi_tlb_flush_range(vma->vm_mm, addr, PAGE_SIZE, PAGE_SIZE); + __flush_tlb_range(vma->vm_mm, addr, PAGE_SIZE, PAGE_SIZE); } =20 void flush_tlb_range(struct vm_area_struct *vma, unsigned long start, unsigned long end) { - __sbi_tlb_flush_range(vma->vm_mm, start, end - start, PAGE_SIZE); + __flush_tlb_range(vma->vm_mm, start, end - start, PAGE_SIZE); } #ifdef CONFIG_TRANSPARENT_HUGEPAGE void flush_pmd_tlb_range(struct vm_area_struct *vma, unsigned long start, unsigned long end) { - __sbi_tlb_flush_range(vma->vm_mm, start, end - start, PMD_SIZE); + __flush_tlb_range(vma->vm_mm, start, end - start, PMD_SIZE); } #endif --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 531ABC3DA7D for ; Tue, 3 Jan 2023 14:14:08 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237889AbjACONi (ORCPT ); Tue, 3 Jan 2023 09:13:38 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59594 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237851AbjACONM (ORCPT ); Tue, 3 Jan 2023 09:13:12 -0500 Received: from mail-pl1-x62c.google.com (mail-pl1-x62c.google.com [IPv6:2607:f8b0:4864:20::62c]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 483FB11452 for ; Tue, 3 Jan 2023 06:13:11 -0800 (PST) Received: by mail-pl1-x62c.google.com with SMTP id y19so13667776plb.2 for ; Tue, 03 Jan 2023 06:13:11 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=0l+i8uOcQCPjKaNrtjuOObll0B7/vooLhOJlMTFIbSo=; b=gBGy2J1w2sNV+FKC+CwdJral87hxtnjAvNKrX/I7Kx9XxFh1Rv51fQJYZ54uQYIS4B 1tNxT5y1i3npT2XnNuRbkMyEQ6UCS60F7ECqvYbv2KKfqlDRWssFgN13KZXsyCnuVW6h ac5+UnnYYm1V6/zMXBm26gN2RzGkXdr9TNei2tXCQozIE6IsXk4lhqtYmMGuHOn3Yc31 OzBdso1OdRcYYWVh6/o8XyArpoMuNJr7DRn1HXlWRCOhHMZLolGVNoF+0SwtbCopui7q FqvYYrBqonmUg1ooGXM9ltL6E42vT+Q+jZm8lFncGX3YoCneec+V8bs0a9pHM6/TcMNe F0SQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=0l+i8uOcQCPjKaNrtjuOObll0B7/vooLhOJlMTFIbSo=; b=B+BkvkZg5OXdNsXlT3Oc3+2PPNcZ91qx2nGIJTrOB9yPQ7bi9mdmO+H92vfCWUkb9Y 5k0k6yvMdKvX2P3iiJKQPeM+B0bmexXRqkB21K/4lUo42nELtpGks4WhcZiNtSxrv/Tw x6qgjYPqLnlDOaHXhF0jtU7n/FUFKatfHYLtssXBY1LGNgXnjaQ5QZVOAW0fmPiY6hbj fOTMB3ufRLApCosWE/36n9XH86rc7zTdMhHHjaw082KU3PFInzmMCxA7feFwIcCbt187 5Kz3BqcX+XFs2PVrAJWvh5JnCyo/NWB1udYd6cJEDoOnlS5CyN0KlYQJ/5C0AgJmscKC bAKw== X-Gm-Message-State: AFqh2kqAR0ECPHNQ1FMEG6vyuYqCKmfoj5O3GlXpiu/69VWtCDhh2lNB /ZPPWvyrD4f9CMgIN4llFImtUw== X-Google-Smtp-Source: AMrXdXs3feqGr1SDeiC+wJOoPhGUm9jY6uwlop6e/pzHdUts4/koTGdcSOhacx6snIxjQ3SET/uLJw== X-Received: by 2002:a17:902:e84c:b0:189:340c:20d2 with SMTP id t12-20020a170902e84c00b00189340c20d2mr71560778plg.23.1672755190683; Tue, 03 Jan 2023 06:13:10 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.13.05 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:13:10 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel , Atish Patra Subject: [PATCH v16 7/9] RISC-V: Use IPIs for remote icache flush when possible Date: Tue, 3 Jan 2023 19:42:19 +0530 Message-Id: <20230103141221.772261-8-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" If we have specialized interrupt controller (such as AIA IMSIC) which allows supervisor mode to directly inject IPIs without any assistance from M-mode or HS-mode then using such specialized interrupt controller, we can do remote icache flushe directly from supervisor mode instead of using the SBI RFENCE calls. This patch extends remote icache flush functions to use supervisor mode IPIs whenever direct supervisor mode IPIs.are supported by interrupt controller. Signed-off-by: Anup Patel Reviewed-by: Atish Patra Acked-by: Palmer Dabbelt --- arch/riscv/mm/cacheflush.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/riscv/mm/cacheflush.c b/arch/riscv/mm/cacheflush.c index 3cc07ed45aeb..b093727494eb 100644 --- a/arch/riscv/mm/cacheflush.c +++ b/arch/riscv/mm/cacheflush.c @@ -19,7 +19,7 @@ void flush_icache_all(void) { local_flush_icache_all(); =20 - if (IS_ENABLED(CONFIG_RISCV_SBI)) + if (IS_ENABLED(CONFIG_RISCV_SBI) && !riscv_use_ipi_for_rfence()) sbi_remote_fence_i(NULL); else on_each_cpu(ipi_remote_fence_i, NULL, 1); @@ -67,7 +67,8 @@ void flush_icache_mm(struct mm_struct *mm, bool local) * with flush_icache_deferred(). */ smp_mb(); - } else if (IS_ENABLED(CONFIG_RISCV_SBI)) { + } else if (IS_ENABLED(CONFIG_RISCV_SBI) && + !riscv_use_ipi_for_rfence()) { sbi_remote_fence_i(&others); } else { on_each_cpu_mask(&others, ipi_remote_fence_i, NULL, 1); --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 63583C53210 for ; Tue, 3 Jan 2023 14:14:08 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237957AbjACONo (ORCPT ); Tue, 3 Jan 2023 09:13:44 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59956 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237887AbjACONc (ORCPT ); Tue, 3 Jan 2023 09:13:32 -0500 Received: from mail-pj1-x102c.google.com (mail-pj1-x102c.google.com [IPv6:2607:f8b0:4864:20::102c]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 8A3251181A for ; Tue, 3 Jan 2023 06:13:16 -0800 (PST) Received: by mail-pj1-x102c.google.com with SMTP id o31-20020a17090a0a2200b00223fedffb30so31209704pjo.3 for ; Tue, 03 Jan 2023 06:13:16 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=fOsF09BSv6WEjfktoanH0vG9zVuY2VCJ6w8ZQR8SIzw=; b=E8TWf5DkaPepHOkkXVQ+z8+CLsbLMAAvnxqIbnjKBZsOqm5DJLf4puRyphTaXyU7A3 O20QSTl/gD6BfMbEjrMBswkQ2h+vdKWGGbkpnphb7jN+WHlAF8wJ9JJYAhz3yGD5s/f2 M4qufJ2rkuyfvuEUfC8iBM8MGVWgvY+NJotvQZHFI7YhFe/5Jv6/ZqFn8lwng1OwLSfA 8S0bJb64BFUYOOnDUHcZ/6iT6d54Q54fwUs5SN3W4qp3X3nnTy+i7Z3IkQucMwzklVhr ahzWYcQ9DvTEWEHqAFDrsnEwT9HyIodujAoltPh0wMNEwmdDz4oe7eQQ2rKO3pAP1FuI 4YBg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=fOsF09BSv6WEjfktoanH0vG9zVuY2VCJ6w8ZQR8SIzw=; b=OTD4Ooez5om3P897jdZhLGYTdZ8exCdtx0keWSY6CTtTe2ml7utBrJecD8hLY4MRLW 4W/UUcjzKmYyfp5UEzj8vCm6Tj2kRpmlXyv4HAVk3iHq9UDj0YSXhUpymsQGesxbdOXP TfNFXnVBWchYJKdx70Ad5p30Pq9e8+wLpFOPGTBV/tjucFApQnB4KAfP58+TZJFZvbwa FscZ7QKVWmWTrmT2sNiWEpOPVIPgeRWllSYJogI5sKUwMXNJLabTp6pxeJ+gsGhSluoe O0JrHjkyuiKsraJ2kE5qYj1nTCZuzVX63wmre+jmCq1WDryfRkzGPScoLJnpZEMNxYz2 VqgQ== X-Gm-Message-State: AFqh2kpRoovGoCV/FGAwF6ubSDi3azGd9Ud17G6Ma7TfZfNuLtee2VtK j5TIEDLIinGCq3CDD7/aqMHLLw== X-Google-Smtp-Source: AMrXdXsuVkKXoZiM/PJQ6X7CsAqviLdEWi7Y0fZ02BkzwCJ5cCR4nkdmJMMRGmSCm2P4swYEtm6ULA== X-Received: by 2002:a17:902:c386:b0:192:709b:9a6a with SMTP id g6-20020a170902c38600b00192709b9a6amr32164806plg.65.1672755196106; Tue, 03 Jan 2023 06:13:16 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.13.10 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:13:15 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel Subject: [PATCH v16 8/9] irqchip/riscv-intc: Add empty irq_eoi() for chained irq handlers Date: Tue, 3 Jan 2023 19:42:20 +0530 Message-Id: <20230103141221.772261-9-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" We add empty irq_eoi() in RISC-V INTC driver for child irqchip drivers (such as PLIC, SBI IPI, CLINT, APLIC, IMSIC, etc) which implement chained handlers for parent per-HART local interrupts. This hels us avoid unnecessary mask/unmask of per-HART local interrupts at the time of handling interrupts. Signed-off-by: Anup Patel Acked-by: Palmer Dabbelt --- drivers/irqchip/irq-riscv-intc.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/drivers/irqchip/irq-riscv-intc.c b/drivers/irqchip/irq-riscv-i= ntc.c index 784d25645704..f229e3e66387 100644 --- a/drivers/irqchip/irq-riscv-intc.c +++ b/drivers/irqchip/irq-riscv-intc.c @@ -46,10 +46,27 @@ static void riscv_intc_irq_unmask(struct irq_data *d) csr_set(CSR_IE, BIT(d->hwirq)); } =20 +static void riscv_intc_irq_eoi(struct irq_data *d) +{ + /* + * The RISC-V INTC driver uses handle_percpu_devid_irq() flow + * for the per-HART local interrupts and child irqchip drivers + * (such as PLIC, SBI IPI, CLINT, APLIC, IMSIC, etc) implement + * chained handlers for the per-HART local interrupts. + * + * In the absence of irq_eoi(), the chained_irq_enter() and + * chained_irq_exit() functions (used by child irqchip drivers) + * will do unnecessary mask/unmask of per-HART local interrupts + * at the time of handling interrupts. To avoid this, we provide + * an empty irq_eoi() callback for RISC-V INTC irqchip. + */ +} + static struct irq_chip riscv_intc_chip =3D { .name =3D "RISC-V INTC", .irq_mask =3D riscv_intc_irq_mask, .irq_unmask =3D riscv_intc_irq_unmask, + .irq_eoi =3D riscv_intc_irq_eoi, }; =20 static int riscv_intc_domain_map(struct irq_domain *d, unsigned int irq, --=20 2.34.1 From nobody Tue Sep 16 12:31:04 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 59A9CC3DA7D for ; Tue, 3 Jan 2023 14:14:40 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237853AbjACOOL (ORCPT ); Tue, 3 Jan 2023 09:14:11 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60102 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237907AbjACONd (ORCPT ); Tue, 3 Jan 2023 09:13:33 -0500 Received: from mail-pj1-x1031.google.com (mail-pj1-x1031.google.com [IPv6:2607:f8b0:4864:20::1031]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4CFE51182B for ; Tue, 3 Jan 2023 06:13:22 -0800 (PST) Received: by mail-pj1-x1031.google.com with SMTP id n65-20020a17090a2cc700b0021bc5ef7a14so31254111pjd.0 for ; Tue, 03 Jan 2023 06:13:22 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=tA+HY8C9FhGSClROg8ZMyX/bN+ae+//jUNHibPXNHtk=; b=e8S/eCAuJpO1vq0OTgX0MCq6k8FTGDeMflkER6JG9mzjFQBwQN+ux9kJ5KvRgCWvCI zmDac8Fykb468GOgbi3eEypP4PfnjYissLl2oscCEN2pKKsLzW50qzmQVgp9FHvH/T+d JoO7TkceGYpYB05CPU4IGVpNvpfgYg5myVJkA1faR6r5LwVDteuF3UyuiB//1Eoii4Lv aDIV3uinCDiy6NhBRijPat60fllKkSiQNaLeglOzEFbTjVwfzfvV49i9TIbu19/k+o2k CCsYlcgYZARsVHXmTpphgjSXo8wjsXSbYKb69+r7/yZ1+ToQMgbGf6HqkGX7VwhrCyGJ qVgA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=tA+HY8C9FhGSClROg8ZMyX/bN+ae+//jUNHibPXNHtk=; b=fm/No/7Lhb2mz4+wPovlVxiEU2XyEiX5Fona36cRHxWULM++tEwdUqYyzR292yAL0w RYx+7EaF+15M6rUf7lu0V0fPk5Rp44kE927CpLZHNPYqLe2XPvjtDGwq38RWm3hNb+fk GlO/b9zfus+iajDylk5Xd9s8Bn4QhglHBloPW9yu7j5FZB+tEGXM0UYy5/6gfiFLCvKm yox65pBh4Zd35LQIfop3XVeOfMYA1w+Cg0ndhADRRGiUvSshMnjmLNuE2ILP1T0tvhxF QK2hz6Uh5pWlTH7S3oaHYLjNFpkTvXMXTD6QVB7YR6uf/PupxSkLrL7BJvBK/uzmiwJQ of4Q== X-Gm-Message-State: AFqh2kp+w30W7vUNkqNgRCKRuJS5+3QlwJN1jyEAC10Re+qbyDt1weuv h8iVwzMAzn3Dk+wNZQVBgG+opg== X-Google-Smtp-Source: AMrXdXttulaHXcvtyTZblGi3ROW8f6QP7p+1BvSxpVJOFGVwEIZH9pfk0MjT85E2Exag8m8UG17POg== X-Received: by 2002:a17:903:251:b0:192:9369:b2fe with SMTP id j17-20020a170903025100b001929369b2femr23780552plh.38.1672755201511; Tue, 03 Jan 2023 06:13:21 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([171.76.85.241]) by smtp.gmail.com with ESMTPSA id x16-20020a1709027c1000b00192b0a07891sm8598286pll.101.2023.01.03.06.13.16 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 03 Jan 2023 06:13:20 -0800 (PST) From: Anup Patel To: Palmer Dabbelt , Paul Walmsley , Thomas Gleixner , Marc Zyngier , Daniel Lezcano Cc: Hector Martin , Sven Peter , Alyssa Rosenzweig , Atish Patra , Alistair Francis , Anup Patel , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, asahi@lists.linux.dev, Anup Patel Subject: [PATCH v16 9/9] irqchip/apple-aic: Move over to core ipi-mux Date: Tue, 3 Jan 2023 19:42:21 +0530 Message-Id: <20230103141221.772261-10-apatel@ventanamicro.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230103141221.772261-1-apatel@ventanamicro.com> References: <20230103141221.772261-1-apatel@ventanamicro.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Marc Zyngier Now that the complexity of the AIC IPI mux has been copied into the core code for the benefit of the riscv architecture, shrink the AIC driver by the same amount by using that infrastructure. Signed-off-by: Marc Zyngier Signed-off-by: Anup Patel Acked-by: Hector Martin Acked-by: Palmer Dabbelt --- drivers/irqchip/Kconfig | 1 + drivers/irqchip/irq-apple-aic.c | 161 ++------------------------------ 2 files changed, 9 insertions(+), 153 deletions(-) diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig index 0e3a7749d87c..9e65345ca3f6 100644 --- a/drivers/irqchip/Kconfig +++ b/drivers/irqchip/Kconfig @@ -659,6 +659,7 @@ config APPLE_AIC bool "Apple Interrupt Controller (AIC)" depends on ARM64 depends on ARCH_APPLE || COMPILE_TEST + select GENERIC_IRQ_IPI_MUX help Support for the Apple Interrupt Controller found on Apple Silicon SoCs, such as the M1. diff --git a/drivers/irqchip/irq-apple-aic.c b/drivers/irqchip/irq-apple-ai= c.c index ae3437f03e6c..eabb3b92965b 100644 --- a/drivers/irqchip/irq-apple-aic.c +++ b/drivers/irqchip/irq-apple-aic.c @@ -292,7 +292,6 @@ struct aic_irq_chip { void __iomem *base; void __iomem *event; struct irq_domain *hw_domain; - struct irq_domain *ipi_domain; struct { cpumask_t aff; } *fiq_aff[AIC_NR_FIQ]; @@ -307,9 +306,6 @@ struct aic_irq_chip { =20 static DEFINE_PER_CPU(uint32_t, aic_fiq_unmasked); =20 -static DEFINE_PER_CPU(atomic_t, aic_vipi_flag); -static DEFINE_PER_CPU(atomic_t, aic_vipi_enable); - static struct aic_irq_chip *aic_irqc; =20 static void aic_handle_ipi(struct pt_regs *regs); @@ -751,98 +747,8 @@ static void aic_ipi_send_fast(int cpu) isb(); } =20 -static void aic_ipi_mask(struct irq_data *d) -{ - u32 irq_bit =3D BIT(irqd_to_hwirq(d)); - - /* No specific ordering requirements needed here. */ - atomic_andnot(irq_bit, this_cpu_ptr(&aic_vipi_enable)); -} - -static void aic_ipi_unmask(struct irq_data *d) -{ - struct aic_irq_chip *ic =3D irq_data_get_irq_chip_data(d); - u32 irq_bit =3D BIT(irqd_to_hwirq(d)); - - atomic_or(irq_bit, this_cpu_ptr(&aic_vipi_enable)); - - /* - * The atomic_or() above must complete before the atomic_read() - * below to avoid racing aic_ipi_send_mask(). - */ - smp_mb__after_atomic(); - - /* - * If a pending vIPI was unmasked, raise a HW IPI to ourselves. - * No barriers needed here since this is a self-IPI. - */ - if (atomic_read(this_cpu_ptr(&aic_vipi_flag)) & irq_bit) { - if (static_branch_likely(&use_fast_ipi)) - aic_ipi_send_fast(smp_processor_id()); - else - aic_ic_write(ic, AIC_IPI_SEND, AIC_IPI_SEND_CPU(smp_processor_id())); - } -} - -static void aic_ipi_send_mask(struct irq_data *d, const struct cpumask *ma= sk) -{ - struct aic_irq_chip *ic =3D irq_data_get_irq_chip_data(d); - u32 irq_bit =3D BIT(irqd_to_hwirq(d)); - u32 send =3D 0; - int cpu; - unsigned long pending; - - for_each_cpu(cpu, mask) { - /* - * This sequence is the mirror of the one in aic_ipi_unmask(); - * see the comment there. Additionally, release semantics - * ensure that the vIPI flag set is ordered after any shared - * memory accesses that precede it. This therefore also pairs - * with the atomic_fetch_andnot in aic_handle_ipi(). - */ - pending =3D atomic_fetch_or_release(irq_bit, per_cpu_ptr(&aic_vipi_flag,= cpu)); - - /* - * The atomic_fetch_or_release() above must complete before the - * atomic_read() below to avoid racing aic_ipi_unmask(). - */ - smp_mb__after_atomic(); - - if (!(pending & irq_bit) && - (atomic_read(per_cpu_ptr(&aic_vipi_enable, cpu)) & irq_bit)) { - if (static_branch_likely(&use_fast_ipi)) - aic_ipi_send_fast(cpu); - else - send |=3D AIC_IPI_SEND_CPU(cpu); - } - } - - /* - * The flag writes must complete before the physical IPI is issued - * to another CPU. This is implied by the control dependency on - * the result of atomic_read_acquire() above, which is itself - * already ordered after the vIPI flag write. - */ - if (send) - aic_ic_write(ic, AIC_IPI_SEND, send); -} - -static struct irq_chip ipi_chip =3D { - .name =3D "AIC-IPI", - .irq_mask =3D aic_ipi_mask, - .irq_unmask =3D aic_ipi_unmask, - .ipi_send_mask =3D aic_ipi_send_mask, -}; - -/* - * IPI IRQ domain - */ - static void aic_handle_ipi(struct pt_regs *regs) { - int i; - unsigned long enabled, firing; - /* * Ack the IPI. We need to order this after the AIC event read, but * that is enforced by normal MMIO ordering guarantees. @@ -857,27 +763,7 @@ static void aic_handle_ipi(struct pt_regs *regs) aic_ic_write(aic_irqc, AIC_IPI_ACK, AIC_IPI_OTHER); } =20 - /* - * The mask read does not need to be ordered. Only we can change - * our own mask anyway, so no races are possible here, as long as - * we are properly in the interrupt handler (which is covered by - * the barrier that is part of the top-level AIC handler's readl()). - */ - enabled =3D atomic_read(this_cpu_ptr(&aic_vipi_enable)); - - /* - * Clear the IPIs we are about to handle. This pairs with the - * atomic_fetch_or_release() in aic_ipi_send_mask(), and needs to be - * ordered after the aic_ic_write() above (to avoid dropping vIPIs) and - * before IPI handling code (to avoid races handling vIPIs before they - * are signaled). The former is taken care of by the release semantics - * of the write portion, while the latter is taken care of by the - * acquire semantics of the read portion. - */ - firing =3D atomic_fetch_andnot(enabled, this_cpu_ptr(&aic_vipi_flag)) & e= nabled; - - for_each_set_bit(i, &firing, AIC_NR_SWIPI) - generic_handle_domain_irq(aic_irqc->ipi_domain, i); + ipi_mux_process(); =20 /* * No ordering needed here; at worst this just changes the timing of @@ -887,55 +773,24 @@ static void aic_handle_ipi(struct pt_regs *regs) aic_ic_write(aic_irqc, AIC_IPI_MASK_CLR, AIC_IPI_OTHER); } =20 -static int aic_ipi_alloc(struct irq_domain *d, unsigned int virq, - unsigned int nr_irqs, void *args) +static void aic_ipi_send_single(unsigned int cpu) { - int i; - - for (i =3D 0; i < nr_irqs; i++) { - irq_set_percpu_devid(virq + i); - irq_domain_set_info(d, virq + i, i, &ipi_chip, d->host_data, - handle_percpu_devid_irq, NULL, NULL); - } - - return 0; -} - -static void aic_ipi_free(struct irq_domain *d, unsigned int virq, unsigned= int nr_irqs) -{ - /* Not freeing IPIs */ + if (static_branch_likely(&use_fast_ipi)) + aic_ipi_send_fast(cpu); + else + aic_ic_write(aic_irqc, AIC_IPI_SEND, AIC_IPI_SEND_CPU(cpu)); } =20 -static const struct irq_domain_ops aic_ipi_domain_ops =3D { - .alloc =3D aic_ipi_alloc, - .free =3D aic_ipi_free, -}; - static int __init aic_init_smp(struct aic_irq_chip *irqc, struct device_no= de *node) { - struct irq_domain *ipi_domain; int base_ipi; =20 - ipi_domain =3D irq_domain_create_linear(irqc->hw_domain->fwnode, AIC_NR_S= WIPI, - &aic_ipi_domain_ops, irqc); - if (WARN_ON(!ipi_domain)) - return -ENODEV; - - ipi_domain->flags |=3D IRQ_DOMAIN_FLAG_IPI_SINGLE; - irq_domain_update_bus_token(ipi_domain, DOMAIN_BUS_IPI); - - base_ipi =3D __irq_domain_alloc_irqs(ipi_domain, -1, AIC_NR_SWIPI, - NUMA_NO_NODE, NULL, false, NULL); - - if (WARN_ON(!base_ipi)) { - irq_domain_remove(ipi_domain); + base_ipi =3D ipi_mux_create(AIC_NR_SWIPI, aic_ipi_send_single); + if (WARN_ON(base_ipi <=3D 0)) return -ENODEV; - } =20 set_smp_ipi_range(base_ipi, AIC_NR_SWIPI); =20 - irqc->ipi_domain =3D ipi_domain; - return 0; } =20 --=20 2.34.1