> No functional change.
>
> Signed-off-by: Thomas Gleixner <tg...@linutronix.de>
Acked-by: David S. Miller <da...@davemloft.net>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Not quite true ... you take out GENERIC_HARDIRQS but select
HAVE_GENERIC_HARDIRQS ... is that right?
> Signed-off-by: Thomas Gleixner <tg...@linutronix.de>
> Cc: Kyle McMartin <ky...@mcmartin.ca>
Plus, it doesn't apply:
Hunk #1 FAILED at 12.
Hunk #2 FAILED at 65.
2 out of 2 hunks FAILED -- saving rejects to file
arch/parisc/Kconfig.rej
It looks like it was made before we did our generic irq conversion in
the 2.6.36 merge window.
James
Nevermind, just saw the rest of the series on linux-arch.
--Kyle
> On Wed, 2011-01-19 at 21:49 +0000, Thomas Gleixner wrote:
> > plain text document attachment (parisc-use-generic-irq-Kconfig.patch)
> > No functional change.
>
> Not quite true ... you take out GENERIC_HARDIRQS but select
> HAVE_GENERIC_HARDIRQS ... is that right?
HAVE_GENERIC_HARDIRQS enables the Kconfig file in kernel/irq/
> > Signed-off-by: Thomas Gleixner <tg...@linutronix.de>
> > Cc: Kyle McMartin <ky...@mcmartin.ca>
>
> Plus, it doesn't apply:
>
> Hunk #1 FAILED at 12.
> Hunk #2 FAILED at 65.
> 2 out of 2 hunks FAILED -- saving rejects to file
> arch/parisc/Kconfig.rej
>
> It looks like it was made before we did our generic irq conversion in
> the 2.6.36 merge window.
You meant 38 window :) Patch is against 38-rc1 but depends on [patch
24/37] genirq: Remove __do_IRQ in this very same series.
Thanks,
tglx
> On Wed, Jan 19, 2011 at 05:17:06PM -0500, Kyle McMartin wrote:
> > I submitted series to clean these 3 up as well (but only got a response
> > from the cris folks...) If the other 2 are effectively unmaintained,
Dammit. Seems they got lost on my side in some mail black hole. I hate
redundant work
> > should I just submit the clean ups to you? (I think there were some
> > build fixes as well, will have to check.)
> >
>
> Nevermind, just saw the rest of the series on linux-arch.
>
> --Kyle
> --
> To unsubscribe from this list: send the line "unsubscribe linux-arch" in
> On Wed, Jan 19, 2011 at 1:47 PM, Thomas Gleixner <tg...@linutronix.de> wrote:
> > Now that rc1 is out, I got reliable information about the remaining
> > users of __do_IRQ(), which block the removal:
> >
> > �arch/cris, arch/h8300, arch/m32r
>
> I'm certainly ok with this series. It would be good to finally get rid
> of this legacy interface, and there's no longer any reason to hold it
> up. And getting rid of hundreds of lines of code is good too.
Will wait a day for desperate screams, then collect the incoming acks
and push it into git for your consumption.
Thanks,
tglx
Acked-by: Tony Luck <tony...@intel.com>
> No functional change.
>
> Signed-off-by: Thomas Gleixner <tg...@linutronix.de>
> Cc: David Howells <dhow...@redhat.com>
Acked-by: David Howells <dhow...@redhat.com>
> All architectures are finally converted. Remove the cruft.
>
> Signed-off-by: Thomas Gleixner <tg...@linutronix.de>
> Cc: Richard Henderson <r...@twiddle.net>
> Cc: Mike Frysinger <vap...@gentoo.org>
> Cc: David Howells <dhow...@redhat.com>
> Cc: Tony Luck <tony...@intel.com>
> Cc: Greg Ungerer <ge...@uclinux.org>
> Cc: Michal Simek <mon...@monstr.eu>
> Cc: David Howells <dhow...@redhat.com>
> Cc: Kyle McMartin <ky...@mcmartin.ca>
> Cc: Benjamin Herrenschmidt <be...@kernel.crashing.org>
> Cc: Chen Liqin <liqin...@sunplusct.com>
> Cc: "David S. Miller" <da...@davemloft.net>
> Cc: Chris Metcalf <cmet...@tilera.com>
> Cc: Jeff Dike <jd...@addtoit.com>
Finally!
Ah, OK, thanks.
> > > Signed-off-by: Thomas Gleixner <tg...@linutronix.de>
> > > Cc: Kyle McMartin <ky...@mcmartin.ca>
> >
> > Plus, it doesn't apply:
> >
> > Hunk #1 FAILED at 12.
> > Hunk #2 FAILED at 65.
> > 2 out of 2 hunks FAILED -- saving rejects to file
> > arch/parisc/Kconfig.rej
> >
> > It looks like it was made before we did our generic irq conversion in
> > the 2.6.36 merge window.
>
> You meant 38 window :)
No ... it went in when Kyle was maintainer ... so that's at least 37 or
before.
> Patch is against 38-rc1 but depends on [patch
> 24/37] genirq: Remove __do_IRQ in this very same series.
Ah, great ... missed that one. With this it all seems to work, so you
can add my ack to the parisc piece.
James
Acked-by: Michal Simek <mon...@monstr.eu>
--
Michal Simek, Ing. (M.Eng)
w: www.monstr.eu p: +42-0-721842854
Maintainer of Linux kernel 2.6 Microblaze Linux - http://www.monstr.eu/fdt/
Microblaze U-BOOT custodian
I'm not sure it's legal to have two "config TILE" blocks in this Kconfig
file; there's already one further down.
Looking at other top-level arch Kconfigs, it does seem common to have the
"config $ARCH" block at the top, though, so if you want to move the
existing "config TILE" block to the top of the Kconfig as part of this
change, that's fine too.
--
Chris Metcalf, Tilera Corp.
http://www.tilera.com
It is OK to have two "config TILE" blocks.
You should take care of moving the other one up after this goes in.
Sam
In that case I'm happy to ack the original patch.
Acked-by: Chris Metcalf <cmet...@tilera.com>
--
Chris Metcalf, Tilera Corp.
http://www.tilera.com
--
Yeah, Amen !
Acked-by: Benjamin Herrenschmidt <be...@kernel.crashing.org>
Cheers,
Ben.
Missed that.
> It is OK to have two "config TILE" blocks.
> You should take care of moving the other one up after this goes in.
Will consolidate before sending it upstream.
Thanks,
tglx
arch/cris, arch/h8300, arch/m32r
The following series converts these architectures to proper flow
handlers and removes __do_IRQ(). Aside of that if cleans up the irq
related Kconfigs in the architectures and switches them to the generic
Kconfig in kernel/irq/
The __do_IRQ() removal is only compile tested, but
- the CRIS maintainer said some time ago he is fine with breakage
and will deal with the fallout
- h8300 and m32r are basically unmaintained. The last updates from
the maintainer side have been more than a year ago.
The conversion of all 3 architectures is straight forward and with low risc,
though it's only compile tested with all available defconfigs as far as it
goes. Some of them do not build in vanilla either, but the converted files
build fine.
The last chunk of this patch series is a straight forward conversion
of architectures to use the generic irq Kconfig file. Compile tested
and boot tested on a few.
That said, if there are no objections and Linus does not go berserk on
me, I'd like to get over with this right now.
Thanks,
tglx
---
Documentation/feature-removal-schedule.txt | 8 -
arch/alpha/Kconfig | 19 --
arch/blackfin/Kconfig | 17 --
arch/cris/Kconfig | 6
arch/cris/arch-v10/kernel/irq.c | 41 +----
arch/cris/arch-v32/kernel/irq.c | 50 +-----
arch/cris/kernel/irq.c | 6
arch/frv/Kconfig | 9 -
arch/h8300/Kconfig | 6
arch/h8300/kernel/irq.c | 50 ++----
arch/ia64/Kconfig | 26 ---
arch/m32r/Kconfig | 11 -
arch/m32r/kernel/irq.c | 10 -
arch/m32r/platforms/m32104ut/setup.c | 58 ++-----
arch/m32r/platforms/m32700ut/setup.c | 214 +++++++++-------------------
arch/m32r/platforms/mappi/setup.c | 78 +++-------
arch/m32r/platforms/mappi2/setup.c | 89 +++--------
arch/m32r/platforms/mappi3/setup.c | 92 ++++--------
arch/m32r/platforms/oaks32r/setup.c | 65 ++------
arch/m32r/platforms/opsput/setup.c | 220 +++++++++--------------------
arch/m32r/platforms/usrv/setup.c | 115 +++++----------
arch/m68knommu/Kconfig | 9 -
arch/microblaze/Kconfig | 11 -
arch/mips/Kconfig | 3
arch/mn10300/Kconfig | 8 -
arch/parisc/Kconfig | 18 --
arch/powerpc/Kconfig | 28 ---
arch/score/Kconfig | 10 -
arch/sparc/Kconfig | 9 -
arch/tile/Kconfig | 19 --
arch/um/Kconfig.common | 6
arch/um/Kconfig.um | 3
include/linux/irqdesc.h | 14 -
kernel/irq/Kconfig | 3
kernel/irq/handle.c | 111 --------------
35 files changed, 413 insertions(+), 1029 deletions(-)
I'm certainly ok with this series. It would be good to finally get rid
of this legacy interface, and there's no longer any reason to hold it
up. And getting rid of hundreds of lines of code is good too.
Linus
I submitted series to clean these 3 up as well (but only got a response
from the cris folks...) If the other 2 are effectively unmaintained,
should I just submit the clean ups to you? (I think there were some
build fixes as well, will have to check.)
--Kyle
Wasn't this supposed to be "select HAVE_GENERIC_HARDIRQS"?
/me studying generic hardirqs...
>
> config AM33_2
> def_bool n
> @@ -76,10 +77,6 @@ config QUICKLIST
> config ARCH_HAS_ILOG2_U32
> def_bool y
>
> -# Use the generic interrupt handling code in kernel/irq/
> -config GENERIC_HARDIRQS
> - def_bool y
> -
> config HOTPLUG_CPU
> def_bool n
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
One could hope this was nicely documented in kernel/irq/Kconfig.
But that file contains only 9 lines of comments for 10 config symbols.
Maybe it is just me but I was struggeling understanding the purpose
of these config symbols.
I found the individual bits of genirq nicely documented in most cases.
But the overall picture was/is hard to get.
And I never got enough understanding to write up something myself :-(
Sam
> On Sat, Mar 05, 2011 at 10:02:33PM +0100, Geert Uytterhoeven wrote:
> > On Wed, Jan 19, 2011 at 22:49, Thomas Gleixner <tg...@linutronix.de> wrote:
> > > No functional change.
> > >
> > > Signed-off-by: Thomas Gleixner <tg...@linutronix.de>
> > > Cc: David Howells <dhow...@redhat.com>
> > > ---
> > > arch/mn10300/Kconfig | 5 +----
> > > 1 file changed, 1 insertion(+), 4 deletions(-)
> > >
> > > Index: linux-2.6/arch/mn10300/Kconfig
> > > ===================================================================
> > > --- linux-2.6.orig/arch/mn10300/Kconfig
> > > +++ linux-2.6/arch/mn10300/Kconfig
> > > @@ -1,6 +1,7 @@
> > > config MN10300
> > > def_bool y
> > > select HAVE_OPROFILE
> > > + select GENERIC_HARDIRQS
> >
> > Wasn't this supposed to be "select HAVE_GENERIC_HARDIRQS"?
Crap, yes.
> > /me studying generic hardirqs...
>
> One could hope this was nicely documented in kernel/irq/Kconfig.
> But that file contains only 9 lines of comments for 10 config symbols.
>
> Maybe it is just me but I was struggeling understanding the purpose
> of these config symbols.
>
> I found the individual bits of genirq nicely documented in most cases.
> But the overall picture was/is hard to get.
>
> And I never got enough understanding to write up something myself :-(
Point taken. /me goes and adds comments :)
tglx