Re: [GIT]: Sparc

Previous thread: [GIT]: Networking fixes by David Miller on Sunday, April 27, 2008 - 7:56 pm. (1 message)

Next thread: Kconfig 'depend' vs. 'select' by David Miller on Sunday, April 27, 2008 - 8:45 pm. (8 messages)
To: <torvalds@...>
Cc: <akpm@...>, <sparclinux@...>, <linux-kernel@...>
Subject: [GIT]: Sparc
Date: Sunday, April 27, 2008 - 8:01 pm

Several fixes and cleanups:

1) Remove SunOS signal frame support, that Linux used for like
a month some 12 years ago :-) I added an assertion that checks
to make sure 32-bit processes request the newer style signal
frames, which I doubt will ever trigger since Jakub Jelinek
and myself are probably the only two people on the planet who
ever had such Linux binaries on their system :-)

2) Revamp sparcspkr probing to use the generic OF platform bus
probing methods, instead of the cruddy ISA and EBUS layers.

This along with deleting the never-enabled ISA floppy support
code allows us to delete the sparc64 specific ISA bus layer
entirely.

Getting rid of the EBUS layer is much hardware, and will require
some coordination with 32-bit sparc.

3) Fix regression in the new "regs->orig_eax == -1" x86 style syscall
restart handling on sparc64. We'd accidently restart a successful
fork/vfork/clone in the child. Doh! :)

4) SBUS video driver fixes and cleanups from Robert Reif.

Please pull, thanks a lot!

The following changes since commit c3bf9bc243092c53946fd6d8ebd6dc2f4e572d48:
Linus Torvalds (1):
Merge branch 'for-linus' of git://git.kernel.org/.../x86/linux-2.6-x86-bigbox-bootmem-v3

are available in the git repository at:

master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6.git master

David S. Miller (11):
sparc64: Fix wedged irq regression.
sparc64: Cleanups and corrections for arch/sparc64/Kconfig
sparc64: Kill CONFIG_SPARC32_COMPAT
sparc64: Kill ISA_FLOPPY_WORKS code.
sparc64: Do not ignore 'pmu' device ranges.
input: Rewrite sparcspkr device probing.
sparc64: Kill unused local ISA bus layer.
sparc64: Kill bogus RT_ALIGNEDSZ macro from signal.c
sparc: Remove old style signal frame support.
sparc64: Clean up handling of pt_regs trap type encoding.
sparc64: Fix accidental syscall restart on child return from clone/fork/vfork.

Robert Rei...

To: David Miller <davem@...>
Cc: <torvalds@...>, <akpm@...>, <sparclinux@...>, <linux-kernel@...>
Date: Monday, April 28, 2008 - 2:29 am

Is a WARN_ON the right check for user-triggerable behaviour? I think

harder? :)

--

To: <hch@...>
Cc: <torvalds@...>, <akpm@...>, <sparclinux@...>, <linux-kernel@...>
Date: Monday, April 28, 2008 - 2:37 am

From: Christoph Hellwig <hch@infradead.org>

I used WARN_ON_ONCE, which I think is sufficient.
--

Previous thread: [GIT]: Networking fixes by David Miller on Sunday, April 27, 2008 - 7:56 pm. (1 message)

Next thread: Kconfig 'depend' vs. 'select' by David Miller on Sunday, April 27, 2008 - 8:45 pm. (8 messages)