Re: Some debugging patches on top of -mm

[Alexander Nyberg]

> This patch, known as page-owner-tracking-leak-detector.patch has
> apparently been sitting in Andrew's *-mm for the last two months.
> 
> I just noticed it now, when reading mm/page_alloc.c.
> 
> I'd like to know if the #ifdef's and CONFIG_PAGE_OWNER specific code
> can be removed from page_alloc.c, and put in a header file.  Ideally,
> you patch would add just one line to the __alloc_pages() code - a call
> to set_page_owner() that either became no code (a static inline empty
> function) or a call to your code, if this feature was CONFIG enabled.
> 
> The *.c files are where all the logic comes together, and it is vital
> to the long term readability of these files that we avoid #ifdef's in
> these files.  Any one feature can be ifdef'd in, with seeming little
> harm to the readability of the code (especially in the eyes of the
> author of that particular bit of ifdef'd code ;).  But imagine what
> a deity-awful mess these files would be if we had all been doing that
> over the years with our various favorite features.
> 

Yes, it was a quick hack when akpm said that he'd like a page tracking
mechanism. I said myself that I thought it was too ugly to go into
mainline. Later I tried to make it some kind of generic framework that
all arches could use but there was no interest.

If you wish to make it nicer/mergeable it's all yours.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/