Re: [PATCH] mm readahead: Fix sys_readahead breakage by reverting 2MB limit (bug 79111)

[John Stoffel]

>>>>> "Linus" == Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> writes:

Linus> On Thu, Jul 3, 2014 at 11:22 AM, Linus Torvalds
Linus> <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>> 
>> So the bugzilla entry worries me a bit - we definitely do not want to
>> regress in case somebody really relied on timing - but without more
>> specific information I still think the real bug is just in the
>> man-page.

Linus> Side note: the 2MB limit may be too small. 2M is peanuts on modern
Linus> machines, even for fairly slow IO, and there are lots of files (like
Linus> glibc etc) that people might want to read-ahead during boot. We
Linus> already do bigger read-ahead if people just do "read()" system calls.
Linus> So I could certainly imagine that we should increase it.

Linus> I do *not* think we should bow down to insane man-pages that have
Linus> always been wrong, though, and I don't think we should increase it to
Linus> "let's just read-ahead a whole ISO image" kind of sizes..

This is one of those perenial questions of how to tune this.  I agree
we should increase the number, but shouldn't it be based on both the
amount of memory in the machine, number of devices (or is it all just
one big pool?) and the speed of the actual device doing readahead?
Doesn't make sense to do 32mb of readahead on a USB 1.1 thumb drive or
even a CDROM.  But maybe it does for USB3 thumb drives?  

John
--
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/