Re: Locking in the (now generic) GPIO infrastructure?

[Mikael Pettersson <mikpe@it.uu.se>]

David Brownell writes:
 > On Wednesday 04 June 2008, Leon Woestenberg wrote:
 > > include/asm-arm/arch-ixp4xx/platform.h:
 > > static inline void gpio_line_set(u8 line, int value)
 > > {
 > >         if (value == IXP4XX_GPIO_HIGH)
 > >             *IXP4XX_GPIO_GPOUTR |= (1 << line);
 > >         else if (value == IXP4XX_GPIO_LOW)
 > >             *IXP4XX_GPIO_GPOUTR &= ~(1 << line);
 > > }
 > > 
 > > Under a Linux kernel where multiple drivers are accessing GPIO, the
 > > latter does not seem safe against preemption (assuming the memory
 > > read-modify-write is not atomic).
 > > 
 > > Shouldn't GPIO access be protected against concurrent access here?
 > 
 > Well, against an IRQ in the middle of those read/modify/write
 > sequences hidden by the "|=" and "&=" syntax.  Last I knew,
 > no XScale CPUs support on-chip SMP.

The IOP342 has two XScale cores on-chip. However, these cores are
still only ARMv5TE and Linux wants ARMv6 or newer for SMP, so I
don't know to what extent the IOP342 is supported by Linux.