Linux-ARM-Kernel Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Joshua Covington <joshuacov@gmail.com>
To: Thomas Gleixner <tglx@kernel.org>
Cc: Florian Fainelli <florian.fainelli@broadcom.com>,
	Radu Rendec <radu@rendec.net>,
	Broadcom internal kernel review list
	<bcm-kernel-feedback-list@broadcom.com>,
	Stanimir Varbanov <svarbanov@suse.de>,
	Andrea della Porta <andrea.porta@suse.com>,
	Pengpeng Hou <pengpeng@iscas.ac.cn>,
	linux-arm-kernel@lists.infradead.org,
	linux-rpi-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org,
	Joshua Covington <joshuacov@gmail.com>
Subject: [PATCH] irqchip/irq-bcm2712-mip: Drop MSI_CHIP_FLAG_SET_ACK
Date: Sat, 21 Nov 2026 16:38:09 +0000	[thread overview]
Message-ID: <20261121163809.86620-1-joshuacov@gmail.com> (raw)

Commit f35508b93a2f ("irqchip/irq-bcm2712-mip: Set EOI/ACK flags in
msi_parent_ops") made the PCI MSI device chips ack through
irq_chip_ack_parent(). Neither the MIP middle chip nor the GIC below it
implements irq_ack, so any ack is a NULL pointer call.

The MIP vectors are handled by handle_fasteoi_irq(), which never acks,
so this only shows when mask_ack_irq() runs, e.g. when a chained
handler is removed. The RP1 driver does that on every teardown since
commit 82aa033fef52 ("misc: rp1: clear chained IRQ handlers on
teardown"), and the kernel panics:

  pc : 0x0
  lr : irq_chip_ack_parent+0x1c/0x28
  Call trace:
   __irq_do_set_handler+0x248/0x24c
   irq_set_chained_handler_and_data+0x6c/0x94
   rp1_unregister_interrupts+0x3c/0xa8

The MIP hierarchy has no use for an ack, like the drivers fixed by
commit fb0ea6e4878a ("irqchip: Drop MSI_CHIP_FLAG_SET_ACK from
unsuspecting MSI drivers").

Fixes: f35508b93a2f ("irqchip/irq-bcm2712-mip: Set EOI/ACK flags in msi_parent_ops")
Signed-off-by: Joshua Covington <joshuacov@gmail.com>
---
The panic was reported on BCM2712 with 6.18.61 and later when the
rp1_pci probe fails (https://github.com/openwrt/openwrt/pull/26530#issuecomment-6975561491).
Tested on a Raspberry Pi CM5 with 6.18.64 and the Raspberry Pi
downstream patches: with this patch, unbinding and binding rp1_pci
again completes without a panic. Applies unchanged to v7.4-rc3 and
builds there without warnings (W=1, GCC 13 and clang 18).

 drivers/irqchip/irq-bcm2712-mip.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/irqchip/irq-bcm2712-mip.c b/drivers/irqchip/irq-bcm2712-mip.c
index 06a6ab627..12d9e3846 100644
--- a/drivers/irqchip/irq-bcm2712-mip.c
+++ b/drivers/irqchip/irq-bcm2712-mip.c
@@ -163,7 +163,7 @@ static const struct irq_domain_ops mip_middle_domain_ops = {
 static const struct msi_parent_ops mip_msi_parent_ops = {
 	.supported_flags	= MIP_MSI_FLAGS_SUPPORTED,
 	.required_flags		= MIP_MSI_FLAGS_REQUIRED,
-	.chip_flags		= MSI_CHIP_FLAG_SET_EOI | MSI_CHIP_FLAG_SET_ACK,
+	.chip_flags		= MSI_CHIP_FLAG_SET_EOI,
 	.bus_select_token       = DOMAIN_BUS_GENERIC_MSI,
 	.bus_select_mask	= MATCH_PCI_MSI,
 	.prefix			= "MIP-MSI-",
-- 
2.43.0



                 reply	other threads:[~2026-11-21 16:38 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20261121163809.86620-1-joshuacov@gmail.com \
    --to=joshuacov@gmail.com \
    --cc=andrea.porta@suse.com \
    --cc=bcm-kernel-feedback-list@broadcom.com \
    --cc=florian.fainelli@broadcom.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rpi-kernel@lists.infradead.org \
    --cc=pengpeng@iscas.ac.cn \
    --cc=radu@rendec.net \
    --cc=svarbanov@suse.de \
    --cc=tglx@kernel.org \
    /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