From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Marc Zyngier <maz@kernel.org>,
Geert Uytterhoeven <geert@linux-m68k.org>,
Sasha Levin <sashal@kernel.org>,
tglx@linutronix.de
Subject: [PATCH AUTOSEL 5.10 21/23] irqchip/gic: Work around broken Renesas integration
Date: Tue, 28 Sep 2021 01:56:42 -0400 [thread overview]
Message-ID: <20210928055645.172544-21-sashal@kernel.org> (raw)
In-Reply-To: <20210928055645.172544-1-sashal@kernel.org>
From: Marc Zyngier <maz@kernel.org>
[ Upstream commit b78f26926b17cc289e4f16b63363abe0aa2e8efc ]
Geert reported that the GIC driver locks up on a Renesas system
since 005c34ae4b44f085 ("irqchip/gic: Atomically update affinity")
fixed the driver to use writeb_relaxed() instead of writel_relaxed().
As it turns out, the interconnect used on this system mandates
32bit wide accesses for all MMIO transactions, even if the GIC
architecture specifically mandates for some registers to be byte
accessible. Gahhh...
Work around the issue by crudly detecting the offending system,
and falling back to an inefficient RMW+lock implementation.
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/CAMuHMdV+Ev47K5NO8XHsanSq5YRMCHn2gWAQyV-q2LpJVy9HiQ@mail.gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/irqchip/irq-gic.c | 52 ++++++++++++++++++++++++++++++++++++++-
1 file changed, 51 insertions(+), 1 deletion(-)
diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c
index 6053245a4754..176f5f06432d 100644
--- a/drivers/irqchip/irq-gic.c
+++ b/drivers/irqchip/irq-gic.c
@@ -107,6 +107,8 @@ static DEFINE_RAW_SPINLOCK(cpu_map_lock);
#endif
+static DEFINE_STATIC_KEY_FALSE(needs_rmw_access);
+
/*
* The GIC mapping of CPU interfaces does not necessarily match
* the logical CPU numbering. Let's use a mapping as returned
@@ -777,6 +779,25 @@ static int gic_pm_init(struct gic_chip_data *gic)
#endif
#ifdef CONFIG_SMP
+static void rmw_writeb(u8 bval, void __iomem *addr)
+{
+ static DEFINE_RAW_SPINLOCK(rmw_lock);
+ unsigned long offset = (unsigned long)addr & 3UL;
+ unsigned long shift = offset * 8;
+ unsigned long flags;
+ u32 val;
+
+ raw_spin_lock_irqsave(&rmw_lock, flags);
+
+ addr -= offset;
+ val = readl_relaxed(addr);
+ val &= ~GENMASK(shift + 7, shift);
+ val |= bval << shift;
+ writel_relaxed(val, addr);
+
+ raw_spin_unlock_irqrestore(&rmw_lock, flags);
+}
+
static int gic_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
bool force)
{
@@ -791,7 +812,10 @@ static int gic_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
if (cpu >= NR_GIC_CPU_IF || cpu >= nr_cpu_ids)
return -EINVAL;
- writeb_relaxed(gic_cpu_map[cpu], reg);
+ if (static_branch_unlikely(&needs_rmw_access))
+ rmw_writeb(gic_cpu_map[cpu], reg);
+ else
+ writeb_relaxed(gic_cpu_map[cpu], reg);
irq_data_update_effective_affinity(d, cpumask_of(cpu));
return IRQ_SET_MASK_OK_DONE;
@@ -1384,6 +1408,30 @@ static bool gic_check_eoimode(struct device_node *node, void __iomem **base)
return true;
}
+static bool gic_enable_rmw_access(void *data)
+{
+ /*
+ * The EMEV2 class of machines has a broken interconnect, and
+ * locks up on accesses that are less than 32bit. So far, only
+ * the affinity setting requires it.
+ */
+ if (of_machine_is_compatible("renesas,emev2")) {
+ static_branch_enable(&needs_rmw_access);
+ return true;
+ }
+
+ return false;
+}
+
+static const struct gic_quirk gic_quirks[] = {
+ {
+ .desc = "broken byte access",
+ .compatible = "arm,pl390",
+ .init = gic_enable_rmw_access,
+ },
+ { },
+};
+
static int gic_of_setup(struct gic_chip_data *gic, struct device_node *node)
{
if (!gic || !node)
@@ -1400,6 +1448,8 @@ static int gic_of_setup(struct gic_chip_data *gic, struct device_node *node)
if (of_property_read_u32(node, "cpu-offset", &gic->percpu_offset))
gic->percpu_offset = 0;
+ gic_enable_of_quirks(node, gic_quirks, gic);
+
return 0;
error:
--
2.33.0
next prev parent reply other threads:[~2021-09-28 5:59 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-28 5:56 [PATCH AUTOSEL 5.10 01/23] spi: rockchip: handle zero length transfers without timing out Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 02/23] platform/x86: touchscreen_dmi: Add info for the Chuwi HiBook (CWI514) tablet Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 03/23] platform/x86: touchscreen_dmi: Update info for the Chuwi Hi10 Plus (CWI527) tablet Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 04/23] nfsd: back channel stuck in SEQ4_STATUS_CB_PATH_DOWN Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 05/23] btrfs: replace BUG_ON() in btrfs_csum_one_bio() with proper error handling Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 06/23] btrfs: fix mount failure due to past and transient device flush error Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 07/23] net: mdio: introduce a shutdown method to mdio device drivers Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 08/23] xen-netback: correct success/error reporting for the SKB-with-fraglist case Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 09/23] sparc64: fix pci_iounmap() when CONFIG_PCI is not set Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 10/23] ext2: fix sleeping in atomic bugs on error Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 11/23] scsi: sd: Free scsi_disk device via put_device() Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 12/23] usb: testusb: Fix for showing the connection speed Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 13/23] usb: dwc2: check return value after calling platform_get_resource() Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 14/23] habanalabs/gaudi: fix LBW RR configuration Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 15/23] selftests: be sure to make khdr before other targets Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 16/23] selftests:kvm: fix get_warnings_count() ignoring fscanf() return warn Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 17/23] nvme-fc: update hardware queues before using them Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 18/23] nvme-fc: avoid race between time out and tear down Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 19/23] thermal/drivers/tsens: Fix wrong check for tzd in irq handlers Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 20/23] scsi: ses: Retry failed Send/Receive Diagnostic commands Sasha Levin
2021-09-28 5:56 ` Sasha Levin [this message]
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 22/23] smb3: correct smb3 ACL security descriptor Sasha Levin
2021-09-28 5:56 ` [PATCH AUTOSEL 5.10 23/23] tools/vm/page-types: remove dependency on opt_file for idle page tracking Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210928055645.172544-21-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=geert@linux-m68k.org \
--cc=linux-kernel@vger.kernel.org \
--cc=maz@kernel.org \
--cc=stable@vger.kernel.org \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox