[Horst von Brand]
> Would tsort(1) perhaps help?
I'm betting Linus would never go for using tsort to resolve such issues
-- unless tsort output is guaranteed to be stable (the docs for GNU
textutils don't say). This would be for the same reason that he
rejected the partial ordering in the LINK_
Keith Owens <[EMAIL PROTECTED]> said:
[...]
> Messing about with conditional compilation because the link order is
> incorrect is the wrong fix. The mtd/Makefile must link the objects in
> the correct order.
>
> cfi_probe.o needs to come after cfi_cmdset_000?.o.
> doc_probe.o needs to come aft
> David Woodhouse <[EMAIL PROTECTED]> wrote:
> >The conditional compilation is far more obvious to people than subtle
> >issues with link order. So I prefer to avoid the latter at all costs.
>
> The rest of the kernel already depends totally on these "subtle" issues
> with link order. Why should
On Sun, 17 Dec 2000, Keith Owens wrote:
> Your choice. Just bear in mind that if CONFIG_MODULES=y but mtd
> objects are built into the kernel then mtd _must_ have a correct link
> order. Consider a config with CONFIG_MODULES=y but every mtd option is
> set to 'y', link order is critical.
Yep,
On Sun, 17 Dec 2000 11:39:50 + (GMT),
David Woodhouse <[EMAIL PROTECTED]> wrote:
>On Sun, 17 Dec 2000, Keith Owens wrote:
>
>> The rest of the kernel already depends totally on these "subtle" issues
>> with link order. Why should mtd be different?
>
>Because I maintain the MTD code and I wan
On Sun, 17 Dec 2000, Keith Owens wrote:
> The rest of the kernel already depends totally on these "subtle" issues
> with link order. Why should mtd be different?
Because I maintain the MTD code and I want it to be. I think the link
order dependencies are ugly, unnecessary and far more likely to
On Sun, 17 Dec 2000 10:44:09 + (GMT),
David Woodhouse <[EMAIL PROTECTED]> wrote:
>The conditional compilation is far more obvious to people than subtle
>issues with link order. So I prefer to avoid the latter at all costs.
The rest of the kernel already depends totally on these "subtle" issu
On Sun, 17 Dec 2000, Keith Owens wrote:
> Messing about with conditional compilation because the link order is
> incorrect is the wrong fix. The mtd/Makefile must link the objects in
> the correct order.
The conditional compilation is far more obvious to people than subtle
issues with link orde
On Sun, 17 Dec 2000 10:01:07 + (GMT),
David Woodhouse <[EMAIL PROTECTED]> wrote:
>On Sun, 17 Dec 2000, Keith Owens wrote:
>
>> Somebody changed include/linux/mtd/map.h between 2.4.0-test11 and
>> test12. That change is wrong, it adds conditional complexity where it
>> is not required - inter
On Sat, 16 Dec 2000, Rasmus Andersen wrote:
> Various files in drivers/mtd references cfi_probe (by way of do_cfi_probe).
> This function is static and thus not shared. The following patch removes
> the static declaration but if it is What Was Intended I do not know. It
> makes the kernel link, h
On Sun, 17 Dec 2000, Keith Owens wrote:
> Somebody changed include/linux/mtd/map.h between 2.4.0-test11 and
> test12. That change is wrong, it adds conditional complexity where it
> is not required - inter_module_xxx works even without CONFIG_MODULES.
> cfi_probe should still be static.
No. Thi
On Sat, 16 Dec 2000 23:07:01 +0100,
Rasmus Andersen <[EMAIL PROTECTED]> wrote:
>Various files in drivers/mtd references cfi_probe (by way of do_cfi_probe).
>This function is static and thus not shared. The following patch removes
>the static declaration but if it is What Was Intended I do not kno
12 matches
Mail list logo
