From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 09B4DC77B61 for ; Sat, 8 Apr 2023 10:41:48 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:MIME-Version:References:In-Reply-To: Subject:Cc:To:From:Message-ID:Date:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=RRCuOAiezbg3jFmc7FSjS+kalV+sLuL2o2Wf/rzMYjQ=; b=KCYU4+0lJCPN2G GXFU14CBBSErBTLZstz0Lno4OG/8r/BRuG0LEG7vqvz19T+AHmtOmapzrfknyY1rh1zpP9s/fZgvO mOILIm0HliZh1kvgo6CTwou+QqGZ6NmDrDRDEWBd2GDRx/1IFXbg4Eu56oOkGt5100641T+KBGGQh 5JSNYHgWMS32QhL4u7fcXH/6Eh06O7lOc8qlYFph31IcQVnwVM9Z3dY0oRXrycPRpN68jpx/YcOfn rlGfwhyRdCOeGD5d1gSv6YPhVorc2nDZZlUKCGeuaFIE0bIDIUMOrOJFmMyK6P+nw33QdJJFGr/EQ EcyExEtbGskKiqUYa+lg==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.96 #2 (Red Hat Linux)) id 1pl60D-00CBYp-1b; Sat, 08 Apr 2023 10:41:01 +0000 Received: from dfw.source.kernel.org ([2604:1380:4641:c500::1]) by bombadil.infradead.org with esmtps (Exim 4.96 #2 (Red Hat Linux)) id 1pl60A-00CBYJ-1m for linux-arm-kernel@lists.infradead.org; Sat, 08 Apr 2023 10:41:00 +0000 Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 1436460B18; Sat, 8 Apr 2023 10:40:58 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 57C31C433EF; Sat, 8 Apr 2023 10:40:57 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1680950457; bh=m8ilFxdsCznzMJ+asDJJkOkNv6yH4u5xRqwg6cI8ZE4=; h=Date:From:To:Cc:Subject:In-Reply-To:References:From; b=U8g6xRLchn8jADVBQJm6+fWpPivvXt1eKuh4MNcZelBE9S5n0YYQAiN1w4e+4w0wQ c63PztmXG4D4Sb8GbHTQFdwSer6ZzO1+eGqSAUcQED20YvU748t+uv5TZx7wtHdeIQ cgl4oh8d9oZz7Zhix4kPpAdo8oruGkJ55awKwyqbTzI6vPv1RDSLEaWXom08swKA2L XpNJKvhguG02K7Dne8SJC60e/tvGzkqS0eJ169D9A7laIE+djbmnXFEo5Pzdv1thwO mj+ufaaK/CkH09WXj0PSPUzYbHZrhLbgsIL49HrliQW5GpUXn0TarQ/ScPxVhJMuNH WF83l0cDGMxjw== Received: from sofa.misterjones.org ([185.219.108.64] helo=wait-a-minute.misterjones.org) by disco-boy.misterjones.org with esmtpsa (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.95) (envelope-from ) id 1pl607-006u5C-33; Sat, 08 Apr 2023 11:40:55 +0100 Date: Sat, 08 Apr 2023 11:40:54 +0100 Message-ID: <87ile664i1.wl-maz@kernel.org> From: Marc Zyngier To: Vignesh Raghavendra Cc: Nishanth Menon , Tero Kristo , Santosh Shilimkar , Thomas Gleixner , , Subject: Re: [PATCH RFC v2 2/2] irqchip: irq-ti-sci-inta: Add direct mapped interrupts In-Reply-To: <20230327-irq-affinity-upstream-v2-2-1474e518f1cb@ti.com> References: <20230327-irq-affinity-upstream-v2-0-1474e518f1cb@ti.com> <20230327-irq-affinity-upstream-v2-2-1474e518f1cb@ti.com> User-Agent: Wanderlust/2.15.9 (Almost Unreal) SEMI-EPG/1.14.7 (Harue) FLIM-LB/1.14.9 (=?UTF-8?B?R29qxY0=?=) APEL-LB/10.8 EasyPG/1.0.0 Emacs/27.1 (x86_64-pc-linux-gnu) MULE/6.0 (HANACHIRUSATO) MIME-Version: 1.0 (generated by SEMI-EPG 1.14.7 - "Harue") X-SA-Exim-Connect-IP: 185.219.108.64 X-SA-Exim-Rcpt-To: vigneshr@ti.com, nm@ti.com, kristo@kernel.org, ssantosh@kernel.org, tglx@linutronix.de, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org X-SA-Exim-Mail-From: maz@kernel.org X-SA-Exim-Scanned: No (on disco-boy.misterjones.org); SAEximRunCond expanded to false X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20230408_034058_668786_005D8189 X-CRM114-Status: GOOD ( 36.73 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org On Mon, 27 Mar 2023 16:04:27 +0100, Vignesh Raghavendra wrote: > > Certain high throughput peripherals such as ethernet etc may need > dedicated IRQ lines so the IRQ load can be balanced across cores using > IRQ affinity controls. Current driver aggregates multiple events/IRQ to > single wired IRQ line/VINT thus making it impossible to migrate events > selectively. > In order to overcome this limitation, reserve a set of VINTs as direct > interrupts and map them to events that are known to generate high IRQ > load. Use SoC specific table to determine such events. > > This allows affinity management at parent irqchip level (GIC or > ti-sci-intr + GIC). > > Signed-off-by: Vignesh Raghavendra > --- > drivers/irqchip/irq-ti-sci-inta.c | 149 +++++++++- > 1 file changed, 147 insertions(+), 2 deletions(-) > > diff --git a/drivers/irqchip/irq-ti-sci-inta.c b/drivers/irqchip/irq-ti-sci-inta.c > index a3a22edbe0f0..db098bd95edc 100644 > --- a/drivers/irqchip/irq-ti-sci-inta.c > +++ b/drivers/irqchip/irq-ti-sci-inta.c > @@ -21,6 +21,7 @@ > #include > #include > #include > +#include > #include > > #define TI_SCI_DEV_ID_MASK 0xffff > @@ -40,6 +41,20 @@ > #define VINT_STATUS_OFFSET 0x18 > #define VINT_STATUS_MASKED_OFFSET 0x20 > > +#define DEV_DMASS0_INTAGGR_0 28 > +#define DEV_DMASS0_PKTDMA_0 30 > + > +/** > + * struct ti_sci_inta_soc_data - Description of SoC specific data > + * @events_list: Pointer to array of events requiring direct IRQ > + * mapping > + * @events_list_size: size of event_list array > + */ > +struct ti_sci_inta_soc_data { > + unsigned int *events_list; > + int events_list_size; > +}; > + > /** > * struct ti_sci_inta_event_desc - Description of an event coming to > * Interrupt Aggregator. This serves > @@ -64,6 +79,7 @@ struct ti_sci_inta_event_desc { > * @events: Array of event descriptors assigned to this vint. > * @parent_virq: Linux IRQ number that gets attached to parent > * @vint_id: TISCI vint ID > + * @reserved: Indicate if this VINT is reserved for direct mapping > */ > struct ti_sci_inta_vint_desc { > struct irq_domain *domain; > @@ -72,6 +88,7 @@ struct ti_sci_inta_vint_desc { > struct ti_sci_inta_event_desc events[MAX_EVENTS_PER_VINT]; > unsigned int parent_virq; > u16 vint_id; > + bool reserved; > }; > > /** > @@ -81,6 +98,8 @@ struct ti_sci_inta_vint_desc { > * @vint: TISCI resource pointer representing IA interrupts. > * @global_event: TISCI resource pointer representing global events. > * @vint_list: List of the vints active in the system > + * @resv_vint_list: List of the direct vints active in the system. > + * @resv_events: Pointer to list of direct events > * @vint_mutex: Mutex to protect vint_list > * @base: Base address of the memory mapped IO registers > * @pdev: Pointer to platform device. > @@ -97,12 +116,15 @@ struct ti_sci_inta_vint_desc { > * identifier to let sysfw know where it has to program the > * Global Event number. > * @num_vints: Number of VINT lines allocated to this IA. > + * @num_resv_vints: Number of VINT lines reserved for direct mapping. > */ > struct ti_sci_inta_irq_domain { > const struct ti_sci_handle *sci; > struct ti_sci_resource *vint; > struct ti_sci_resource *global_event; > struct list_head vint_list; > + struct list_head resv_vint_list; > + unsigned int *resv_events; > /* Mutex to protect vint list */ > struct mutex vint_mutex; > void __iomem *base; > @@ -112,6 +134,7 @@ struct ti_sci_inta_irq_domain { > int unmapped_cnt; > u16 *unmapped_dev_ids; > int num_vints; > + int num_resv_vints; > }; > > #define to_vint_desc(e, i) container_of(e, struct ti_sci_inta_vint_desc, \ > @@ -265,7 +288,11 @@ static int ti_sci_inta_alloc_parent_irqs(struct irq_domain *domain) > } > vint_desc->parent_virq = parent_virq; > > - list_add_tail(&vint_desc->list, &inta->vint_list); > + if (i < inta->num_resv_vints) > + list_add_tail(&vint_desc->list, &inta->resv_vint_list); > + else > + list_add_tail(&vint_desc->list, &inta->vint_list); > + > irq_set_chained_handler_and_data(vint_desc->parent_virq, > ti_sci_inta_irq_handler, vint_desc); > } > @@ -390,8 +417,58 @@ static void ti_sci_inta_free_irq(struct ti_sci_inta_event_desc *event_desc, > ti_sci_release_resource(inta->global_event, event_desc->global_event); > event_desc->global_event = TI_SCI_RESOURCE_NULL; > event_desc->hwirq = 0; > + vint_desc->reserved = false; > +} > + > +/** > + * ti_sci_inta_alloc_direct_irq - Allocate non-aggregated events/IRQs > + * @data: Pointer to corresponding irq_data > + * > + * Returns allocated event_desc, ERR_PTR otherwise. > + */ > +static struct ti_sci_inta_event_desc *ti_sci_inta_alloc_direct_irq(struct irq_data *data) > +{ > + struct ti_sci_inta_irq_domain *inta = data->domain->host_data; > + struct ti_sci_inta_event_desc *event_desc = NULL; > + struct ti_sci_inta_vint_desc *vint_desc; > + > + mutex_lock(&inta->vint_mutex); > + list_for_each_entry(vint_desc, &inta->resv_vint_list, list) { > + if (vint_desc->reserved) > + continue; > + > + vint_desc->reserved = true; > + event_desc = ti_sci_inta_alloc_event(vint_desc, 0, data->hwirq); > + if (IS_ERR(event_desc)) > + goto unlock; > > + data->parent_data = irq_get_irq_data(vint_desc->parent_virq); > + break; > + } > +unlock: > mutex_unlock(&inta->vint_mutex); > + > + return event_desc; > +} > + > +/** > + * ti_sci_inta_direct_irq - Allocate non-aggregated events/IRQs > + * @data: Pointer to corresponding irq_data > + * > + * Returns true if this IRQ is to be direct mapped on given platform, > + * false otherwise. > + */ > +static bool ti_sci_inta_direct_irq(struct irq_data *data) > +{ > + struct ti_sci_inta_irq_domain *inta = data->domain->host_data; > + int i; > + > + for (i = 0; i < inta->num_resv_vints; i++) { > + if (data->hwirq == inta->resv_events[i]) > + return true; > + } > + > + return false; > } > > /** > @@ -408,7 +485,11 @@ static int ti_sci_inta_request_resources(struct irq_data *data) > { > struct ti_sci_inta_event_desc *event_desc; > > - event_desc = ti_sci_inta_alloc_irq(data->domain, data->hwirq); > + if (ti_sci_inta_direct_irq(data)) > + event_desc = ti_sci_inta_alloc_direct_irq(data); > + else > + event_desc = ti_sci_inta_alloc_irq(data->domain, data->hwirq); > + > if (IS_ERR(event_desc)) > return PTR_ERR(event_desc); > > @@ -485,6 +566,15 @@ static void ti_sci_inta_ack_irq(struct irq_data *data) > static int ti_sci_inta_set_affinity(struct irq_data *d, > const struct cpumask *mask_val, bool force) > { > + int ret; > + > + if (d->parent_data) { > + ret = irq_chip_set_affinity_parent(d, mask_val, force); > + irq_data_update_effective_affinity(d, mask_val); > + > + return ret; > + } > + > return -EINVAL; > } > > @@ -627,9 +717,54 @@ static int ti_sci_inta_get_unmapped_sources(struct ti_sci_inta_irq_domain *inta) > return 0; > } > > +static unsigned int ti_sci_inta_direct_events_am62x[] = { > + /* CPSW etherenti DMA events */ > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4627), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4635), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4643), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4651), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4659), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4667), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4675), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 4683), > + TO_HWIRQ(DEV_DMASS0_PKTDMA_0, 5651), > +}; > + > +static struct ti_sci_inta_soc_data soc_data_am62x = { > + .events_list = ti_sci_inta_direct_events_am62x, > + .events_list_size = ARRAY_SIZE(ti_sci_inta_direct_events_am62x), > +}; I don't think these tables belong in a driver, and they are bound to grow without any obvious limits. You have firmware tables that can express these things. Surely they can be put to a good use. M. -- Without deviation from the norm, progress is not possible. _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel