linux-kernel - [PATCH v3 0/6] PCI/ACPI: Fix firmware first error recovery with root port in reset

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:	Thu,  6 Jun 2013 12:10:45 -0600
From:	Betty Dall <betty.dall@...com>
To:	rjw@...k.pl, bhelgaas@...gle.com, gong.chen@...ux.intel.com,
	greg.pearson@...com
Cc:	ying.huang@...el.com, linux-acpi@...r.kernel.org,
	linux-kernel@...r.kernel.org, linux-pci@...r.kernel.org,
	Betty Dall <betty.dall@...com>
Subject: [PATCH v3 0/6] PCI/ACPI: Fix firmware first error recovery with root port in reset

This patch set fixes a bug on platforms that use firmware first AER.
Firmware can leave the root port in Secondary Bus Reset (SBR) and
communicate this to the OS through the "reset" bit in the flags field
of the HEST table and associated CPER records. Firmware wants to do this
so that the error is contained and the hardware is in a known state.

Without these patches, the root port stays in SBR and the device drivers
cannot recover. These patches recognize when the firmware first root port
is in SBR and bring the root port out of SBR so the devices under the root
port can recover.

The changes have been tested on systems with firmware first that set the
"reset" bit by injecting various hardware errors. The errors successfully
recover.

Changes since v1:
Fixed a typo in the comment of patch 2.
Removed incorrect setting of reset bit in patch 3.

Changes since v2: 
The v2 patch 1/3 was re-written by Bjorn Helgaas and is now patches 1/6
through 3/6.
The v2 patch 2/3 is now 5/6 and changed to directly use the AER_FATAL define
and introduced patch 4/6 to move the defines to a public header file.
The v2 patch 3/3 is now 6/6 and uses the same default reset link function for
both Downstream Ports and Root Ports.

Signed-off-by: Betty Dall <betty.dall@...com>
---
Betty Dall (6):
  PCI/AER: Don't parse HEST table for non-PCIe devices
  PCI/AER: Factor out HEST device type matching
  PCI/AER: Set dev->__aer_firmware_first only for matching devices
  PCI/ACPI: Move AER severity defines to aer.h
  ACPI/APEI: Force fatal AER severity when bus has been reset
  PCI/AER: Provide reset_link for firmware first root port
---
 drivers/acpi/apei/ghes.c           |   10 +++++++
 drivers/pci/pcie/aer/aerdrv.h      |    4 ---
 drivers/pci/pcie/aer/aerdrv_acpi.c |   47 ++++++++++++++++++-----------------
 drivers/pci/pcie/aer/aerdrv_core.c |   17 +++++++------
 include/linux/aer.h                |   16 +++++++----
 5 files changed, 53 insertions(+), 41 deletions(-)

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/