Index
Home
About
Blog
Newsgroups: fa.linux.kernel
From: Linus Torvalds <torvalds@osdl.org>
Subject: Re: [PATCH] fix platform_rename_gsi related ia32 build breakage
Original-Message-ID: <Pine.LNX.4.58.0411111552030.2301@ppc970.osdl.org>
Date: Fri, 12 Nov 2004 00:21:37 GMT
Message-ID: <fa.gstnue0.aiur9k@ifi.uio.no>
On Thu, 11 Nov 2004, Len Brown wrote:
>
> The _gsi in platform_rename_gsi was consistent with the surrounding use
> in the ACPI case. I decided to re-use the same funtion for the MPS case
> for simplicity, even though io_apic.c uses _irq. If you like, I can add
> a synonym using an inline for _irq, but I thought we were moving away
> from using _irq, not towards it.
We _definitely_ prefer "irq" over something else that means the same
thing.
If GSI means some _specific_ interrupt, and thus has additional meaning
over "irq", then by all means, use it, but spell it out. "Global System
Interrupt" means _nothing_ to me. What makes it "global"? What makes it
"system"?
The _only_ thing that uses "gsi" is the MP table stuff, and that's
apparently just from the documentation.
In other words, if it's a normal interrupt, it's "irq" or "interrupt". The
same way a "disk" is a "disk" - it's not a DASD.
Stupid acronyms that don't actually mean anything more than the standard
name are nothing but stupid.
Interrupts are interrupts. We call them something else only if they are
_specific_ interrupts (ie a "NMI" clearly has a very _specific_ meaning,
as has "SCI", although the latter is already obscure enough that it's
probably a good idea to spell it out a bit if it is ever used outside of a
context where use is obvious).
Linus
Newsgroups: fa.linux.kernel
From: Linus Torvalds <torvalds@osdl.org>
Subject: Re: [PATCH] fix platform_rename_gsi related ia32 build breakage
Original-Message-ID: <Pine.LNX.4.58.0411111903460.2301@ppc970.osdl.org>
Date: Fri, 12 Nov 2004 03:16:25 GMT
Message-ID: <fa.guu5t61.8icq1l@ifi.uio.no>
On Thu, 11 Nov 2004, Len Brown wrote:
>
> I agree 100%, and submit that the term "IRQ" fits this definition of a
> "stupid acronym".
There's a huge difference between an acronyn that is well-established, and
one that is _totally_ made up, has no history, and is only used on one
platform.
"irq" is a very traditional shorthand for "interrupt request", and anybody
who has _ever_ worked with any OS on _any_ platform knows _exactly_ what
it is.
In contrast, gsi has _zero_ meaning outside some small ACPI group.
Trust me. Do a poll.
The same way we don't call disks DASD devices do we not call interrupts
gsi's.
And that "we" is not a "royal we". It's a f*cking established _fact_.
Type "irq" into google, and what's the first hit? In fact, EVERY SINGLE
hit on the first page is relevant.
In contrast, type "gsi" into google, and NOT A SINGLE ONE has _anything_
to do with interrupts. Not the first one, not the tenth one.
So stop being idiotic.
Linus
Index
Home
About
Blog