[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <f5tby75ifujq2ka3ku76ezuzar4i7ok7a7etygygdpt2k6n4ar@wful3braajua>
Date: Fri, 1 Aug 2025 06:35:19 -0700
From: Breno Leitao <leitao@...ian.org>
To: Bjorn Helgaas <helgaas@...nel.org>, pandoh@...gle.com
Cc: linux-pci@...r.kernel.org,
Karolina Stolarek <karolina.stolarek@...cle.com>, Weinan Liu <wnliu@...gle.com>,
Martin Petersen <martin.petersen@...cle.com>, Ben Fuller <ben.fuller@...cle.com>,
Drew Walton <drewwalton@...rosoft.com>, Anil Agrawal <anilagrawal@...a.com>,
Tony Luck <tony.luck@...el.com>, Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>,
Sathyanarayanan Kuppuswamy <sathyanarayanan.kuppuswamy@...ux.intel.com>, Lukas Wunner <lukas@...ner.de>,
Jonathan Cameron <Jonathan.Cameron@...wei.com>, Sargun Dhillon <sargun@...a.com>,
"Paul E . McKenney" <paulmck@...nel.org>, Mahesh J Salgaonkar <mahesh@...ux.ibm.com>,
Oliver O'Halloran <oohall@...il.com>, Kai-Heng Feng <kaihengf@...dia.com>,
Keith Busch <kbusch@...nel.org>, Robert Richter <rrichter@....com>,
Terry Bowman <terry.bowman@....com>, Shiju Jose <shiju.jose@...wei.com>,
Dave Jiang <dave.jiang@...el.com>, linux-kernel@...r.kernel.org, linuxppc-dev@...ts.ozlabs.org,
Bjorn Helgaas <bhelgaas@...gle.com>, kernel-team@...a.com, gustavold@...il.com
Subject: Re: [PATCH v8 18/20] PCI/AER: Ratelimit correctable and non-fatal
error logging
On Fri, Aug 01, 2025 at 06:16:29AM -0700, Breno Leitao wrote:
> Hello Jon, Bjorn,
>
> On Thu, May 22, 2025 at 06:21:24PM -0500, Bjorn Helgaas wrote:
> > @@ -790,6 +818,9 @@ void pci_print_aer(struct pci_dev *dev, int aer_severity,
> > trace_aer_event(pci_name(dev), (status & ~mask),
> > aer_severity, tlp_header_valid, &aer->header_log);
> >
> > + if (!aer_ratelimit(dev, info.severity))
> > + return;
>
> I am seeing a kernel NULL pointer in the aer_ratelimit(), where
> dev->aer_info is NULL. This is happening on linus final 6.16 commit id.
Upon closer examination of the code, it appears we can replicate the
functionality of `pci_dev_aer_stats_incr()`, which is similarly invoked
within this code path.
commit 1b4ef90e8397eaf2bc4d0f8a2127d2d75c7ff5e0
Author: Breno Leitao <leitao@...ian.org>
Date: Fri Aug 1 06:32:26 2025 -0700
PCI/AER: Check for NULL aer_info before ratelimiting in pci_print_aer()
Similarly to pci_dev_aer_stats_incr(), pci_print_aer() may be called
when dev->aer_info is NULL. Add a NULL check before proceeding to avoid
calling aer_ratelimit() with a NULL aer_info pointer, returning 1, which
does not rate limit, given this is fatal.
This prevents a kernel crash triggered by dereferencing a NULL pointer
in aer_ratelimit(), ensuring safer handling of PCI devices that lack
AER info. This change aligns pci_print_aer() with pci_dev_aer_stats_incr()
which already performs this NULL check.
Signed-off-by: Breno Leitao <leitao@...ian.org>
Fixes: a57f2bfb4a5863 ("PCI/AER: Ratelimit correctable and non-fatal
error logging")
diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
index 70ac661883672..b5f96fde4dcda 100644
--- a/drivers/pci/pcie/aer.c
+++ b/drivers/pci/pcie/aer.c
@@ -786,6 +786,9 @@ static void pci_rootport_aer_stats_incr(struct pci_dev *pdev,
static int aer_ratelimit(struct pci_dev *dev, unsigned int severity)
{
+ if (!dev->aer_info)
+ return 1;
+
switch (severity) {
case AER_NONFATAL:
return __ratelimit(&dev->aer_info->nonfatal_ratelimit);
Powered by blists - more mailing lists