[PATCH 0/3] IRQ stack support for ARM

[Russell King - ARM Linux admin]

On Wed, Oct 21, 2020 at 01:58:21PM +0200, Arnd Bergmann wrote:
> (replying to my own mail, apparently my normal outgoing email server is
> blacklisted, so resending from @kernel.org)
> 
> On Fri, Oct 16, 2020 at 12:09 PM Arnd Bergmann <arnd at arndb.de> wrote:
> >
> > On Thu, Oct 8, 2020 at 10:32 AM Russell King - ARM Linux admin
> > <linux at armlinux.org.uk> wrote:
> > > On Thu, Oct 08, 2020 at 12:45:30PM +0530, Maninder Singh wrote:
> > > > Observed Stack Overflow on 8KB kernel stack on ARM specially
> > > > incase on network interrupts, which results in undeterministic behaviour.
> > > > So there is need for per cpu dedicated IRQ stack for ARM.
> > > >
> > > > As ARm does not have extra co-processor register
> > > > to save thread info pointer, IRQ stack will be at some
> > > > performance cost, so code is under CONFIG_IRQ_STACK.
> > > >
> > > > and we don't have much knowledge and set up for CLANG
> > > > and ARM_UNWIND, so dependency added for both cases.
> > > >
> > > > Tested patch set with QEMU for latest kernel
> > > > and 4.1 kernel for ARM target with same patch set.
> > >
> > > You need to investigate and show where and why this is happening. My
> > > guess is you have a network driver that uses a lot of kernel stack
> > > space, which itself would be a bug.
> >
> > Agreed.
> >
> > > Note that there are compiler versions out there that mis-optimise and
> > > eat stack space - the kernel build should be warning if a function
> > > uses a large amount of stack.
> >
> > Some more ideas for figuring it out:
> >
> > CONFIG_DEBUG_STACK_USAGE may also be helpful in identifying
> > code paths that are deeply nested with multiple functions taking a
> > lot of stack space, but each one staying under the limit.
> >
> > CONFIG_DEBUG_STACKOVERFLOW would also help here but
> > is not supported on Arm at the moment. There was a patch[1] from
> > Uwe Kleine-König to add this, and I suppose we should still add
> > that, in particular if it helps debug this problem.
> >
> > CONFIG_VMAP_STACK is probably the best way to debug
> > random runtime stack overflows because using a guard page
> > turns random memory corruption into an immediate oops,
> > but I don't think there is an implementation for Arm yet and
> > using a lot of vmalloc space means we might not be able to
> > default to this.
> >
> > Regardless of identifying and fixing the bug Maninder found, I
> > also think that supporting separate async stacks on Arm is useful
> > for determinism. Most of the popular architectures use irqstack
> > for this reason, and I was actually surprised that we don't do it
> > on arch/arm/.
> >
> >      Arnd
> >
> > [1] https://lore.kernel.org/linux-arm-kernel/20200108082913.29710-1-u.kleine-koenig@pengutronix.de/

We don't do it because we don't have a separate register to be able
to store the thread_info pointer, and copying that lump between the
SVC and IRQ stack will add massively to IRQ latency, especially for
older machines.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!