Message-Id: <1408466769-20004-1-git-send-email-daniel.thompson@linaro.org>
Date:	Tue, 19 Aug 2014 17:45:50 +0100
From:	Daniel Thompson <daniel.thompson@...aro.org>
To:	Russell King <linux@....linux.org.uk>
Cc:	Daniel Thompson <daniel.thompson@...aro.org>,
	linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
	kgdb-bugreport@...ts.sourceforge.net, patches@...aro.org,
	linaro-kernel@...ts.linaro.org,
	John Stultz <john.stultz@...aro.org>,
	Anton Vorontsov <anton.vorontsov@...aro.org>,
	Colin Cross <ccross@...roid.com>, kernel-team@...roid.com,
	Rob Herring <robherring2@...il.com>,
	Linus Walleij <linus.walleij@...aro.org>,
	Ben Dooks <ben.dooks@...ethink.co.uk>,
	Catalin Marinas <catalin.marinas@....com>,
	Dave Martin <Dave.Martin@....com>,
	Fabio Estevam <festevam@...il.com>,
	Frederic Weisbecker <fweisbec@...il.com>,
	Nicolas Pitre <nico@...aro.org>
Subject: [PATCH v10 00/19] arm: KGDB NMI/FIQ support

This patchset makes it possible to use kgdb's NMI infrastructure on ARM
platforms.

The patches are seperated into three distinct groups:

1. arm specific changes; these provide multi-platform support for FIQ
   (including raising an IPI using FIQ to ensure effective SMP support)
   and extend ARM KGDB support to use the features provided.

2. irqchip changes; updates to the gic and vic drivers to provide
   support for routing certain interrupt sources to FIQ.

3. serial changes; driver support to allow the UART interrupt to be
   routed to FIQ. The already mainlined kgdb NMI infrastructure (mostly
   found in drivers/tty/serial/kgdb_nmi.c) will re-route the kgdb
   console UART's interrupt signal from IRQ to FIQ. Naturally the UART
   will no longer function normally and will instead be managed by kgdb
   using the polled I/O functions. Any character delivered to the UART
   causes the kgdb handler function to be called.

Tested on qemu (versatile), STiH416 (B2020 board) and Freescale i.MX6
quad (wandboard).

Changes since v9:

- Split PL011 code movement into a seperate patch (Peter Hurley)
- Use container_of() to convert pointers to uart_amba_port (Peter
  Hurley)
- Clean up redundant code from pl011_poll_init (Peter Hurley)
- Call do_unexp_fiq() if we receive a FIQ and CONFIG_FIQ is not set.
- Ensure irq-gic.c does not call FIQ functions unless CONFIG_FIQ is set.
- Introduced patch to avoid ttyNMI0 being enabled by default
  (replaces architecture dependant enable/disable logic).
- Fixed bug in kgdb_fiq_enable_nmi() that causes SysRq-g (non-NMI
  debugging) to wedge the debugger.

Changes since v8:

- Significant rework to separate the FIQ exception handler from kgdb.
  This allows other features typically implemented using NMI on x86
  to reuse the same exception handling code (Russell King)
- Reunited arch/arm and driver code into a single patch series again.
- Added support for raising IPIs using FIQ (makes kgdb quiesce must
  more robust).
- Introduced a workaround for GICv1 devices to avoid FIQs being
  spuriously handled as IRQs.

Changes since v7:

- Introduced ack_fiq() to complement eoi_fiq(). Without this it is
  not possible to meet the GIC specification (previous versions worked
  when tested but are unpredictable according to the specification).
  ack_fiq() also makes is possible for drivers for devices with multiple
  interrupt lines to discover the interrupt source correctly.

Changes since v6:

- Corrected off-by-one comparison in has_fiq() (Nicolas Pitre)
- Rewrote the FIQ stack initialization (Nicolas Pitre). This fixes a
  serious data corruption bug due to bad stack mismanagement.
- Introduced __fiq_abt to ensure lr_abt and spsr_abt are saved and
  restored if we fast-interrupt an abort (Russell King).
- Significantly improved the commenting of the exception handlers.
- Added a call to trace_hardirqs_on() if we clear the I bit as we
  exit the exception handler.

Changes since v5:

- Separated anything not strictly impacting arch/arm.
- Fixed a spurious add/remove of code within the series (there was
  broken code in intermediate patches)

For previous changes see:
http://thread.gmane.org/gmane.linux.ports.arm.kernel/333901

Daniel Thompson (18):
  arm: fiq: Add callbacks to manage FIQ routings
  arm: fiq: Allow ACK and EOI to be passed to the intc
  arm: fiq: Replace default FIQ handler
  arm: smp: Introduce a special IPI signalled using FIQ
  arm: KGDB/KDB FIQ support
  irqchip: gic: Provide support for interrupt grouping
  irqchip: gic: Add support for FIQ management
  irqchip: gic: Remove spin locks from eoi_irq
  irqchip: gic: Add support for IPI FIQ
  irqchip: gic: Group 0 workaround.
  irqchip: vic: Add support for FIQ management
  serial: kgdb_nmi: No CON_ENABLED by default
  serial: amba-pl011: Use container_of() to get uart_amba_port
  serial: amba-pl011: Move pl011_hwinit()
  serial: amba-pl011: Pass FIQ information to KGDB.
  serial: asc: Add support for KGDB's FIQ/NMI mode
  serial: asc: Adopt readl_/writel_relaxed()
  serial: imx: Add support for KGDB's FIQ/NMI mode

Dirk Behme (1):
  serial: imx: clean up imx_poll_get_char()

 arch/arm/Kconfig                |   2 +
 arch/arm/Kconfig.debug          |  19 +++
 arch/arm/include/asm/fiq.h      |  18 +++
 arch/arm/include/asm/hardirq.h  |   2 +-
 arch/arm/include/asm/kgdb.h     |   4 +
 arch/arm/include/asm/smp.h      |   7 +
 arch/arm/kernel/entry-armv.S    | 129 +++++++++++++++--
 arch/arm/kernel/fiq.c           | 139 +++++++++++++++++-
 arch/arm/kernel/kgdb.c          | 117 ++++++++++++++-
 arch/arm/kernel/setup.c         |   8 +-
 arch/arm/kernel/smp.c           |  15 ++
 arch/arm/mach-versatile/core.c  |   2 +-
 drivers/irqchip/irq-gic.c       | 309 +++++++++++++++++++++++++++++++++++++---
 drivers/irqchip/irq-vic.c       |  92 +++++++++---
 drivers/tty/serial/Kconfig      |   2 +-
 drivers/tty/serial/amba-pl011.c | 143 +++++++++++--------
 drivers/tty/serial/imx.c        |  88 +++++++-----
 drivers/tty/serial/kgdb_nmi.c   |   5 +-
 drivers/tty/serial/st-asc.c     |  27 +++-
 include/linux/irqchip/arm-vic.h |   6 +-
 20 files changed, 977 insertions(+), 157 deletions(-)

--
1.9.3

--
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/

Powered by blists - more mailing lists