From nobody Thu Dec 18 08:54:18 2025 Received: from mail-pj1-f43.google.com (mail-pj1-f43.google.com [209.85.216.43]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 0264D38F91 for ; Sun, 9 Feb 2025 04:18:23 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.43 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1739074705; cv=none; b=nRdX+l1c1FlEY0HIhfNPRMqG+LR4Kr66ikcNSPWa7n/Cd44TeI3FN1ME4nN2l8ylqHQ2RVuU0qpRZus+ZZp4znF6Ad3HcW6+KCDS10EmyY1syGdIfuCuXFKPA1iEsr0mJRtjBabMuprMWWtVbLucU1S61LIq28YXgUCwum4OUWI= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1739074705; c=relaxed/simple; bh=/aguk4ZYg6WQI6RlClB+Z+ge32iMkXIXY6eYPLNpAn0=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=qv6unWAY1KoEh6Fvb/MWzhlfXbWFb1rni7UWJN/vmojzInmCZb+GR7eo5Kzpql7slgof2fhmGoFScaacxulltCBFNDEaYYDDpof3xf7bpQMIyYf0wO+i2N2i9FgqohNXjD1s7cQkVjlGH30L+PZSsSjoEB8sXsyvMjsomDr2Ajg= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=ventanamicro.com; spf=pass smtp.mailfrom=ventanamicro.com; dkim=pass (2048-bit key) header.d=ventanamicro.com header.i=@ventanamicro.com header.b=POHjq9fK; arc=none smtp.client-ip=209.85.216.43 Authentication-Results: smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=ventanamicro.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=ventanamicro.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=ventanamicro.com header.i=@ventanamicro.com header.b="POHjq9fK" Received: by mail-pj1-f43.google.com with SMTP id 98e67ed59e1d1-2fa2c1ab145so3276108a91.3 for ; Sat, 08 Feb 2025 20:18:23 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ventanamicro.com; s=google; t=1739074703; x=1739679503; darn=vger.kernel.org; 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=li309BKFytEFNs0s+UiLLaDTQ3aEtuOmqKESJ/Lkwhg=; b=POHjq9fKo8NiN4JECMIL+N+WfJogJ5PLaEc/11vcOxmtM7n/SYmWbzDXLxZrQ99cGe Xv8NkaeFSc29ig2R55+pyVTqf2VBSArdWWbuhM3UcAX/MOOiFc6ATVdJqtJlwPSZl2eA RV0WbUss7ApmgBsdJk8SJWInyeiUyo/qOKAAqcXZ5dm2whVTesyFsGXDyyFOX7BvRMN2 QYtoLdKtPtRJE4Qa+TNyaukyvkux3K8Zx0a9RqpxMaw7xIYn4aoooe5VdugQourDSbs9 e9JoaAuhxlUYJUy9eKTBpir72/9Y9fKZNGkeK224B0vQtkPa5jbGcUqo8DGmeFIWjLYE O0ow== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1739074703; x=1739679503; 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=li309BKFytEFNs0s+UiLLaDTQ3aEtuOmqKESJ/Lkwhg=; b=ZhbN5dCfVkRlWdlbaykPAsQUnC9kMf+/oQEKB8kY7G5+Fjdy9BSmgrUVcQwQmN3/zj FP9svpryB0zs6Q/6ogF+4t2VA9v+ZC71dktQbrwWgYwMFWu8H4zVmMeKfWdqS+SuKi+E FCX6qajPQWnB+bKhQ/IB8t75k0ahpVEj5xWw48EnzgUsJc11hvlf3Cpku/2Nq5LZS9fu zdoAbvmdZAuz9HxNMFa+gT48KqVW9IinXOgQN6X5zh74HGvR4uo/eveQGBYDkcbX2hIo SOGqyMIRlPLhGCnqgvPpAxuoYPhKUF63yo4DxE+PivvYo6Gw373N7t4IePrElTDjY/+l D4Og== X-Forwarded-Encrypted: i=1; AJvYcCWN3ZLPlb7wmmCJKti0hCJ7fBDIgsGSAOWD6f+W363w3oN21FrLomgBoPjNLDQzgd/qux/Tt8Vnv8/+KLU=@vger.kernel.org X-Gm-Message-State: AOJu0YwWF1JrVxARIPrgyM3Q+9b6kFcwzJ7irded4hgwrhrxf3LHDd0X GWSuVUkcKVgBT5i6/Wm3/2DiWNhhz/vSKAKmLjwftJgv3KIRAy8nSn7N4YE/fH8= X-Gm-Gg: ASbGncuwAhmeM0j6/jIXzWg/Tl2tMNQTvs8lrHO6xOB1gOOKobu65eXqUqeQ7PSPDc9 /FRV9+jhqgD6XzksxUW1dFPXRdnuz2GlF2qSWKe7slD82Y0jjZQeBYlKiC8I9XtN2GBFsIC3Y5K yRr7Va5NaCJ0sV6/VDq1586nx5UTKJVGWsW0zl0/wHzKhetbi8y9amzOdzc/k1aFRqeYCUTsMua Q/n98V4zLghM6CVnsp7yda4oW6qkznu+8FKB33/FV5Lry59DCwDauszTaR+zXPULVGXggiHXGs8 6WUunfqXQiIbu9wSFPynD1U1ej9KJterECR1rZKJWGk259PJv1BSH1A= X-Google-Smtp-Source: AGHT+IGr8DIP/za6PXLubagaC8HrCh4Chcf3SqtLF6bJLforSFiwf8+xl00IhhTWFSgMBoJHAPJZQw== X-Received: by 2002:a05:6a20:6f91:b0:1dc:2365:4d7d with SMTP id adf61e73a8af0-1ee03a4744fmr14841962637.16.1739074703102; Sat, 08 Feb 2025 20:18:23 -0800 (PST) Received: from anup-ubuntu-vm.localdomain ([103.97.166.196]) by smtp.gmail.com with ESMTPSA id d2e1a72fcca58-73073eba116sm1898410b3a.124.2025.02.08.20.18.15 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Sat, 08 Feb 2025 20:18:22 -0800 (PST) From: Anup Patel To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , Dave Hansen , x86@kernel.org Cc: hpa@zytor.com, Marc Zyngier , Shawn Guo , Sascha Hauer , Pengutronix Kernel Team , Andrew Lunn , Gregory Clement , Sebastian Hesselbarth , Palmer Dabbelt , Paul Walmsley , Atish Patra , Andrew Jones , Sunil V L , Anup Patel , linux-riscv@lists.infradead.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, imx@lists.linux.dev, Anup Patel Subject: [PATCH v5 08/11] irqchip/riscv-imsic: Separate next and previous pointers in IMSIC vector Date: Sun, 9 Feb 2025 09:46:52 +0530 Message-ID: <20250209041655.331470-9-apatel@ventanamicro.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20250209041655.331470-1-apatel@ventanamicro.com> References: <20250209041655.331470-1-apatel@ventanamicro.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Currently, there is only one "move" pointer in the struct imsic_vector so during vector movement the old vector points to the new vector and new vector points to itself. To support force cleanup of old vector, add separate "move_next" and "move_prev" pointers in the struct imsic_vector where during vector movement the "move_next" pointer of the old vector points to the new vector and the "move_prev" pointer of the new vector points to the old vector. Both "move_next" and "move_prev" pointers are cleared separately by __imsic_local_sync() with a restriction that "move_prev" on the new CPU is cleared only after old CPU has cleared "move_next". Signed-off-by: Anup Patel --- drivers/irqchip/irq-riscv-imsic-early.c | 8 ++- drivers/irqchip/irq-riscv-imsic-state.c | 96 +++++++++++++++++-------- drivers/irqchip/irq-riscv-imsic-state.h | 7 +- 3 files changed, 78 insertions(+), 33 deletions(-) diff --git a/drivers/irqchip/irq-riscv-imsic-early.c b/drivers/irqchip/irq-= riscv-imsic-early.c index c5c2e6929a2f..b5def6268936 100644 --- a/drivers/irqchip/irq-riscv-imsic-early.c +++ b/drivers/irqchip/irq-riscv-imsic-early.c @@ -77,6 +77,12 @@ static void imsic_handle_irq(struct irq_desc *desc) struct imsic_vector *vec; unsigned long local_id; =20 + /* + * Process pending local synchronization instead of waiting + * for per-CPU local timer to expire. + */ + imsic_local_sync_all(false); + chained_irq_enter(chip, desc); =20 while ((local_id =3D csr_swap(CSR_TOPEI, 0))) { @@ -120,7 +126,7 @@ static int imsic_starting_cpu(unsigned int cpu) * Interrupts identities might have been enabled/disabled while * this CPU was not running so sync-up local enable/disable state. */ - imsic_local_sync_all(); + imsic_local_sync_all(true); =20 /* Enable local interrupt delivery */ imsic_local_delivery(true); diff --git a/drivers/irqchip/irq-riscv-imsic-state.c b/drivers/irqchip/irq-= riscv-imsic-state.c index b97e6cd89ed7..96e994443fc7 100644 --- a/drivers/irqchip/irq-riscv-imsic-state.c +++ b/drivers/irqchip/irq-riscv-imsic-state.c @@ -124,10 +124,11 @@ void __imsic_eix_update(unsigned long base_id, unsign= ed long num_id, bool pend, } } =20 -static void __imsic_local_sync(struct imsic_local_priv *lpriv) +static bool __imsic_local_sync(struct imsic_local_priv *lpriv) { struct imsic_local_config *mlocal; struct imsic_vector *vec, *mvec; + bool ret =3D true; int i; =20 lockdep_assert_held(&lpriv->lock); @@ -143,35 +144,75 @@ static void __imsic_local_sync(struct imsic_local_pri= v *lpriv) __imsic_id_clear_enable(i); =20 /* - * If the ID was being moved to a new ID on some other CPU - * then we can get a MSI during the movement so check the - * ID pending bit and re-trigger the new ID on other CPU - * using MMIO write. + * Clear the pervious vector pointer of the new vector only + * after the movement is complete on the old CPU. */ - mvec =3D READ_ONCE(vec->move); - WRITE_ONCE(vec->move, NULL); - if (mvec && mvec !=3D vec) { + mvec =3D READ_ONCE(vec->move_prev); + if (mvec) { + /* + * If the old vector has not been updated then + * try again in the next sync-up call. + */ + if (READ_ONCE(mvec->move_next)) { + ret =3D false; + continue; + } + + WRITE_ONCE(vec->move_prev, NULL); + } + + /* + * If a vector was being moved to a new vector on some other + * CPU then we can get a MSI during the movement so check the + * ID pending bit and re-trigger the new ID on other CPU using + * MMIO write. + */ + mvec =3D READ_ONCE(vec->move_next); + if (mvec) { if (__imsic_id_read_clear_pending(i)) { mlocal =3D per_cpu_ptr(imsic->global.local, mvec->cpu); writel_relaxed(mvec->local_id, mlocal->msi_va); } =20 + WRITE_ONCE(vec->move_next, NULL); imsic_vector_free(&lpriv->vectors[i]); } =20 skip: bitmap_clear(lpriv->dirty_bitmap, i, 1); } + + return ret; } =20 -void imsic_local_sync_all(void) +#ifdef CONFIG_SMP +static void __imsic_local_timer_start(struct imsic_local_priv *lpriv) +{ + lockdep_assert_held(&lpriv->lock); + + if (!timer_pending(&lpriv->timer)) { + lpriv->timer.expires =3D jiffies + 1; + add_timer_on(&lpriv->timer, smp_processor_id()); + } +} +#else +static inline void __imsic_local_timer_start(struct imsic_local_priv *lpri= v) +{ +} +#endif + +void imsic_local_sync_all(bool force_all) { struct imsic_local_priv *lpriv =3D this_cpu_ptr(imsic->lpriv); unsigned long flags; =20 raw_spin_lock_irqsave(&lpriv->lock, flags); - bitmap_fill(lpriv->dirty_bitmap, imsic->global.nr_ids + 1); - __imsic_local_sync(lpriv); + + if (force_all) + bitmap_fill(lpriv->dirty_bitmap, imsic->global.nr_ids + 1); + if (!__imsic_local_sync(lpriv)) + __imsic_local_timer_start(lpriv); + raw_spin_unlock_irqrestore(&lpriv->lock, flags); } =20 @@ -190,12 +231,7 @@ void imsic_local_delivery(bool enable) #ifdef CONFIG_SMP static void imsic_local_timer_callback(struct timer_list *timer) { - struct imsic_local_priv *lpriv =3D this_cpu_ptr(imsic->lpriv); - unsigned long flags; - - raw_spin_lock_irqsave(&lpriv->lock, flags); - __imsic_local_sync(lpriv); - raw_spin_unlock_irqrestore(&lpriv->lock, flags); + imsic_local_sync_all(false); } =20 static void __imsic_remote_sync(struct imsic_local_priv *lpriv, unsigned i= nt cpu) @@ -216,14 +252,11 @@ static void __imsic_remote_sync(struct imsic_local_pr= iv *lpriv, unsigned int cpu */ if (cpu_online(cpu)) { if (cpu =3D=3D smp_processor_id()) { - __imsic_local_sync(lpriv); - return; + if (__imsic_local_sync(lpriv)) + return; } =20 - if (!timer_pending(&lpriv->timer)) { - lpriv->timer.expires =3D jiffies + 1; - add_timer_on(&lpriv->timer, cpu); - } + __imsic_local_timer_start(lpriv); } } #else @@ -278,8 +311,9 @@ void imsic_vector_unmask(struct imsic_vector *vec) raw_spin_unlock(&lpriv->lock); } =20 -static bool imsic_vector_move_update(struct imsic_local_priv *lpriv, struc= t imsic_vector *vec, - bool new_enable, struct imsic_vector *new_move) +static bool imsic_vector_move_update(struct imsic_local_priv *lpriv, + struct imsic_vector *vec, bool is_old_vec, + bool new_enable, struct imsic_vector *move_vec) { unsigned long flags; bool enabled; @@ -289,7 +323,10 @@ static bool imsic_vector_move_update(struct imsic_loca= l_priv *lpriv, struct imsi /* Update enable and move details */ enabled =3D READ_ONCE(vec->enable); WRITE_ONCE(vec->enable, new_enable); - WRITE_ONCE(vec->move, new_move); + if (is_old_vec) + WRITE_ONCE(vec->move_next, move_vec); + else + WRITE_ONCE(vec->move_prev, move_vec); =20 /* Mark the vector as dirty and synchronize */ bitmap_set(lpriv->dirty_bitmap, vec->local_id, 1); @@ -322,8 +359,8 @@ void imsic_vector_move(struct imsic_vector *old_vec, st= ruct imsic_vector *new_ve * interrupt on the old vector while device was being moved * to the new vector. */ - enabled =3D imsic_vector_move_update(old_lpriv, old_vec, false, new_vec); - imsic_vector_move_update(new_lpriv, new_vec, enabled, new_vec); + enabled =3D imsic_vector_move_update(old_lpriv, old_vec, true, false, new= _vec); + imsic_vector_move_update(new_lpriv, new_vec, false, enabled, old_vec); } =20 #ifdef CONFIG_GENERIC_IRQ_DEBUGFS @@ -386,7 +423,8 @@ struct imsic_vector *imsic_vector_alloc(unsigned int hw= irq, const struct cpumask vec =3D &lpriv->vectors[local_id]; vec->hwirq =3D hwirq; vec->enable =3D false; - vec->move =3D NULL; + vec->move_next =3D NULL; + vec->move_prev =3D NULL; =20 return vec; } diff --git a/drivers/irqchip/irq-riscv-imsic-state.h b/drivers/irqchip/irq-= riscv-imsic-state.h index 391e44280827..f02842b84ed5 100644 --- a/drivers/irqchip/irq-riscv-imsic-state.h +++ b/drivers/irqchip/irq-riscv-imsic-state.h @@ -23,7 +23,8 @@ struct imsic_vector { unsigned int hwirq; /* Details accessed using local lock held */ bool enable; - struct imsic_vector *move; + struct imsic_vector *move_next; + struct imsic_vector *move_prev; }; =20 struct imsic_local_priv { @@ -74,7 +75,7 @@ static inline void __imsic_id_clear_enable(unsigned long = id) __imsic_eix_update(id, 1, false, false); } =20 -void imsic_local_sync_all(void); +void imsic_local_sync_all(bool force_all); void imsic_local_delivery(bool enable); =20 void imsic_vector_mask(struct imsic_vector *vec); @@ -87,7 +88,7 @@ static inline bool imsic_vector_isenabled(struct imsic_ve= ctor *vec) =20 static inline struct imsic_vector *imsic_vector_get_move(struct imsic_vect= or *vec) { - return READ_ONCE(vec->move); + return READ_ONCE(vec->move_prev); } =20 void imsic_vector_move(struct imsic_vector *old_vec, struct imsic_vector *= new_vec); --=20 2.43.0