Ralf Baechle wrote:
> On Thu, Jun 14, 2007 at 03:54:47PM -0600, Marc St-Jean wrote:
>
> General comment all the first ~ 2900 of your patch is that it's rather
> heavy on #ifdef. #ifdef is a nasty construct that has a tendency to hard
> to read code which in trun results in bugs. Or a test compile fails to
> find real issues in the code because it's hidden in a dead #if construct.
I don't disagree but I do not have the resources to completely rewrite all
code using #ifdef at this time. The best we can do is to ensure to eliminate
them as we update the core platform support.
> > diff --git a/include/asm-mips/pmc-sierra/msp71xx/msp_regops.h
> b/include/asm-mips/pmc-sierra/msp71xx/msp_regops.h
> > new file mode 100644
> > index 0000000..60a5a38
> > --- /dev/null
> > +++ b/include/asm-mips/pmc-sierra/msp71xx/msp_regops.h
> > @@ -0,0 +1,236 @@
> > +/*
> > + * SMP/VPE-safe functions to access "registers" (see note).
> > + *
> > + * NOTES:
> > +* - These macros use ll/sc instructions, so it is your
> responsibility to
> > + * ensure these are available on your platform before including this
> file.
> > + * - The MIPS32 spec states that ll/sc results are undefined for
> uncached
> > + * accesses. This means they can't be used on HW registers accessed
> > + * through kseg1. Code which requires these macros for this purpose
> must
> > + * front-end the registers with cached memory "registers" and have a
> single
> > + * thread update the actual HW registers.
>
> You basically betting on undefined behaviour of the architecture. I did
> indeed verify this specific case with a processor design guy and the answer
> was a "should be ok". That is I heared people speak in a more convincing
> tone already ;-)
>
> A SC with an uncached address on some other MIPS processors will always
> fail.
> So this isn't just a theoretical footnote in a manual.
>
> The way things are implemented LL/SC I am certain that uncached LL/SC will
> fail on any MIPS multiprocessor system. Fortunately while SMTC pretends
> to be multiprocessor it's really just a single core, which saves your day.
I added this comment after this was discussed previously in the list, I have
updated the first part to be more clear about the limitations:
* SMTC/VPE-safe functions to access "registers" (IMPORTANT: see NOTES).
*
* Copyright 2005-2007 PMC-Sierra, Inc.
*
* NOTES:
* - Some of the macros defined in this file use LL/SC instructions which
* are architecture specific. There use on MSP71XX is dependant on the
* behavior of the MIPS 34k core in these devices. They will not work on
* other multi-processor architectures or possibly on future VPE-based
* cores.
At that time I also eliminated all use of macros containing LL/SC on uncached
memory outside of the core architecture specific (arch/mips/pmc-sierra/msp71xx)
code. The only driver currently using macros with LL/SC is the TWI LED driver
and it's to access a cached shared memory interface.
> > + * - A maximum of 2k of code can be inserted between ll and sc. Every
> > + * memory accesses between the instructions will increase the chance of
> > + * sc failing and having to loop.
>
> Any memory access between LL/SC makes the LL/SC sequence invalid that is
> it will have undefined effects.
Also a comment added after last discussion. Could you please expand on what
you mean by "undefined"? Are you saying the SC will not report the access?
I thought the reason these instructions exist was to detect an access near
the memory location referenced by them.
Updated to read:
* - A maximum of 2k of code can be inserted between ll and sc. Every
* memory accesses between the instructions will increase the chance of
* sc failing and having to loop. Only perform the necessary logical
* operations on register values in between these two instructions.
> > + * - There is a bug on the R10000 chips which has a workaround. If you
> > + * are affected by this bug, make sure to define the symbol
> 'R10000_LLSC_WAR'
> > + * to be non-zero. If you are using this header from within linux,
> you may
> > + * include <asm/war.h> before including this file to have this defined
> > + * appropriately for you.
>
> > +#ifndef __ASM_REGOPS_H__
> > +#define __ASM_REGOPS_H__
> > +
> > +#include <linux/types.h>
> > +
> > +#include <asm/war.h>
> > +
> > +#ifndef R10000_LLSC_WAR
> > +#define R10000_LLSC_WAR 0
> > +#endif
>
> This symbol is supposed to be defined by <asm/war.h> only. Anyway, this
> #ifndef will never be true because you already include <asm/war.h>, so
> this is dead code.
Thanks, I have removed. This file had originally been written to live in
include/asm-mips.
> > +#if R10000_LLSC_WAR == 1
> > +#define __beqz "beqzl "
> > +#else
> > +#define __beqz "beqz "
> > +#endif
> > +
> > +#ifndef _LINUX_TYPES_H
> > +typedef unsigned int u32;
> > +#endif
>
> Redefining a stanard Linux type is a no-no as is relying on include
> wrapper symbols like _LINUX_TYPES_H. Anyway, this #ifndef will never
> be true because you already include <linux/types.h>, so this is dead code.
Removed, similar situation as above.
> > +static inline u32 read_reg32(volatile u32 *const addr,
> > + u32 const mask)
> > +{
> > + u32 temp;
> > +
> > + __asm__ __volatile__(
> > + " .set push \n"
> > + " .set noreorder \n"
> > + " lw %0, %1 # read \n"
> > + " and %0, %2 # mask \n"
> > + " .set pop \n"
> > + : "=&r" (temp)
> > + : "m" (*addr), "ir" (mask));
> > +
> > + return temp;
> > +}
>
> No need for inline assembler here; plain C can achieve the same. Or just
> use a standard Linux function such as readl() or ioread32() or similar.
OK, used C.
> > +/*
> > + * For special strange cases only:
> > + *
> > + * If you need custom processing within a ll/sc loop, use the
> following macros
> > + * VERY CAREFULLY:
> > + *
> > + * u32 tmp; <-- Define a variable
> to hold the data
> > + *
> > + * custom_read_reg32(address, tmp); <-- Reads the address
> and put the value
> > + * in the 'tmp' variable
> given
> > + *
> > + * From here on out, you are (basicly) atomic, so don't do
> anything too
> > + * fancy!
> > + * Also, this code may loop if the end of this block fails to write
> > + * everything back safely due do the other CPU, so do NOT do anything
> > + * with side-effects!
> > + *
> > + * custom_write_reg32(address, tmp); <-- Writes back 'tmp'
> safely.
> > + */
> > +#define custom_read_reg32(address, tmp) \
> > + __asm__ __volatile__( \
> > + " .set push \n" \
> > + " .set mips3 \n" \
> > + "1: ll %0, %1 #custom_read_reg32 \n" \
> > + " .set pop \n" \
> > + : "=r" (tmp), "=m" (*address) \
> > + : "m" (*address))
> > +
> > +#define custom_write_reg32(address, tmp) \
> > + __asm__ __volatile__( \
> > + " .set push \n" \
> > + " .set mips3 \n" \
> > + " sc %0, %1 #custom_write_reg32 \n" \
> > + " "__beqz"%0, 1b \n" \
> > + " nop \n" \
> > + " .set pop \n" \
> > + : "=&r" (tmp), "=m" (*address) \
> > + : "0" (tmp), "m" (*address))
>
> These two are *really* fragile stuff. Modern gcc rearranges code in
> amazing ways, so you might end up with other loads or stores being moved
> into the ll/sc sequence or the 1: label of another inline assembler
> construct being taken as the destination of the branch. So I would
> suggest to safely store the two function in a nice yellow barrel ;-)
I'm probably missing some European cultural nuance, unless of course
you mean toxic waste :) Dropped.
> General suggestion, you can make about every access atomic if you do
> something like
>
> #include <linux/modules.h>
> #include <linux/spinlocks.h>
>
> DEFINE_SPINLOCK(register_lock);
> EXPORT_SYMBOL(register_lock);
>
> static inline void set_value_reg32(u32 *const addr,
> u32 const mask,
> u32 const value)
> {
> unsigned long flags;
> u32 bits;
>
> spinlock_irqsave(®ister_lock, flags);
> bits = readl(addr);
> bits &= mask;
> bits |= value;
> writel(bits, addr);
> }
>
> Maybe slower but definately more portable and not waiting before some
> CPU designer screws your code by accident :-)
Yes but code running on the RTOS can't pull in spinlock source from linux.
Marc
|