]> git.hungrycats.org Git - linux/log
linux
22 years agoMerge delerium.codemonkey.org.uk:/mnt/nfs/neologic/bar/src/kernel/2.6/trees/bk-linus
Dave Jones [Wed, 2 Jun 2004 02:46:03 +0000 (03:46 +0100)]
Merge delerium.codemonkey.org.uk:/mnt/nfs/neologic/bar/src/kernel/2.6/trees/bk-linus
into delerium.codemonkey.org.uk:/mnt/nfs/neologic/bar/src/kernel/2.6/trees/cpufreq

22 years ago[CPUFREQ] Fix build of longhaul.
Dave Jones [Tue, 1 Jun 2004 14:49:42 +0000 (15:49 +0100)]
[CPUFREQ] Fix build of longhaul.
Arjan noted that in some cases, the build fails.
This should fix it up.

22 years ago[PATCH] Avoid excessive stack usage in NFS
Peter Osterlund [Mon, 31 May 2004 08:11:45 +0000 (01:11 -0700)]
[PATCH] Avoid excessive stack usage in NFS

Assigning '*wdata' with a dynamic initializer creates a
temporary structure copy on the stack, and then the final
data is initialized with a "memcpy()".

As a result, these NFS functions use more than 800 bytes
of stack-space.

Changing the code to just do a memset followed by explicit
initialization of the non-zero member variables takes the
stack usage down to 36 bytes.

Here is a patch that does exactly that.

22 years agoppc64: more user address fixups.
Linus Torvalds [Mon, 31 May 2004 05:35:54 +0000 (22:35 -0700)]
ppc64: more user address fixups.

Removed one (harmless) direct user pointer access.

22 years agoppc64: add more user annotations to ptrace.c
Linus Torvalds [Mon, 31 May 2004 05:21:49 +0000 (22:21 -0700)]
ppc64: add more user annotations to ptrace.c

22 years agoppc64: fix more user pointers in signal handling
Linus Torvalds [Mon, 31 May 2004 05:17:16 +0000 (22:17 -0700)]
ppc64: fix more user pointers in signal handling

22 years agosparse: make x86 user pointer checks stricter.
Linus Torvalds [Mon, 31 May 2004 04:19:40 +0000 (21:19 -0700)]
sparse: make x86 user pointer checks stricter.

Also use "__force" when forcing the user address to "unsigned long"
for the address limit check.

22 years agoppc64: missing __user annotations noticed by stricter checks.
Linus Torvalds [Mon, 31 May 2004 03:41:08 +0000 (20:41 -0700)]
ppc64: missing __user annotations noticed by stricter checks.

22 years agoppc64: check more of the user access functions for proper arguments.
Linus Torvalds [Mon, 31 May 2004 03:39:30 +0000 (20:39 -0700)]
ppc64: check more of the user access functions for proper arguments.

22 years agosparse: add "__force" type attribute for forcing a cast.
Linus Torvalds [Mon, 31 May 2004 03:38:48 +0000 (20:38 -0700)]
sparse: add "__force" type attribute for forcing a cast.

22 years ago[PATCH] sparse: asus_acpi dereference of userland pointers
Alexander Viro [Mon, 31 May 2004 02:59:55 +0000 (19:59 -0700)]
[PATCH] sparse: asus_acpi dereference of userland pointers

Direct sscanf() on userland buffer in procfs ->write_proc().  Fixed.

22 years ago[PATCH] sparse: reiserfs annotation
Alexander Viro [Sun, 30 May 2004 17:38:17 +0000 (10:38 -0700)]
[PATCH] sparse: reiserfs annotation

Everything straigthforward and takes care of almost all sparse warnings
in fs/reiserfs/*.

22 years ago[PATCH] trivial annotation for arch/i386/kernel/*
Alexander Viro [Sun, 30 May 2004 16:17:04 +0000 (09:17 -0700)]
[PATCH] trivial annotation for arch/i386/kernel/*

Several places in arch/i386/kernel were still un-annotated - they didn't
trip copy_.._user(), so they stayed alive.  Caught and killed.

That's not all - there's a big cluster of them in vm86.c, but I haven't
looked into that code yet.

22 years agoppc64: mark the "regshere" marker with proper type information.
Linus Torvalds [Sun, 30 May 2004 15:00:08 +0000 (08:00 -0700)]
ppc64: mark the "regshere" marker with proper type information.

22 years ago[PATCH] sparse alpha: the rest of it
Alexander Viro [Sun, 30 May 2004 14:35:38 +0000 (07:35 -0700)]
[PATCH] sparse alpha: the rest of it

A bunch of "return in void function" dealt with, removed bogus extern
from definition of __load_new_mm_context(), sanitized fscking ugly
CROSS_64K() helper (even if sparse doesn't segfault on that anymore,
it's _still_ fscking ugly).

22 years ago[PATCH] sparse alpha: utimes()
Alexander Viro [Sun, 30 May 2004 14:34:09 +0000 (07:34 -0700)]
[PATCH] sparse alpha: utimes()

... and that's where I'd found do_utimes() crap before I went
grepping ;-)

22 years ago[PATCH] sparse alpha: trivial parts of __user annotation
Alexander Viro [Sun, 30 May 2004 14:33:58 +0000 (07:33 -0700)]
[PATCH] sparse alpha: trivial parts of __user annotation

Trivial stuff in arch/alpha annotated; in two cases of
osf_proplist_syscall() dereferencing userland pointers replaced with
get_user

22 years ago[PATCH] sparse alpha: #if abuses
Alexander Viro [Sun, 30 May 2004 14:33:47 +0000 (07:33 -0700)]
[PATCH] sparse alpha: #if abuses

#if where #ifdef should've been

22 years ago[PATCH] sparse alpha: long constants
Alexander Viro [Sun, 30 May 2004 14:33:36 +0000 (07:33 -0700)]
[PATCH] sparse alpha: long constants

Long constants marked as such

22 years ago[PATCH] sparse alpha: beginning of __user annotation
Alexander Viro [Sun, 30 May 2004 14:33:26 +0000 (07:33 -0700)]
[PATCH] sparse alpha: beginning of __user annotation

Copying primitives annotated.

22 years agoppc64: more explicitly typed constants
Linus Torvalds [Sun, 30 May 2004 14:33:10 +0000 (07:33 -0700)]
ppc64: more explicitly typed constants

22 years agoMerge bk://kernel.bkbits.net/davem/net-2.6
Linus Torvalds [Sun, 30 May 2004 13:31:17 +0000 (06:31 -0700)]
Merge bk://kernel.bkbits.net/davem/net-2.6
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years agoMerge bk://kernel.bkbits.net/davem/sparc-2.6
Linus Torvalds [Sun, 30 May 2004 13:30:21 +0000 (06:30 -0700)]
Merge bk://kernel.bkbits.net/davem/sparc-2.6
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years agoMerge bk://bk.arm.linux.org.uk/linux-2.6-rmk
Linus Torvalds [Sun, 30 May 2004 13:28:41 +0000 (06:28 -0700)]
Merge bk://bk.arm.linux.org.uk/linux-2.6-rmk
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years ago[ARM] Add two more missing __user annotations for sparse.
Russell King [Sun, 30 May 2004 15:40:37 +0000 (16:40 +0100)]
[ARM] Add two more missing __user annotations for sparse.

22 years ago[PATCH] sparse: SIOCGIFCONF handling - the rest of it
Alexander Viro [Sun, 30 May 2004 13:09:42 +0000 (06:09 -0700)]
[PATCH] sparse: SIOCGIFCONF handling - the rest of it

Fixed the type of SIOCGIFCONF callback; inet instance was already
correctly annotated, decnet one was _not_.  Moreover, decnet callback
needed fixing - dereferencing userland address.

22 years ago[PATCH] sparse: tty_driver ->write_proc()
Alexander Viro [Sun, 30 May 2004 13:09:31 +0000 (06:09 -0700)]
[PATCH] sparse: tty_driver ->write_proc()

Made ->write_proc in tty_driver suitable for procfs write callback

22 years ago[PATCH] sparse: n_tty annotation
Alexander Viro [Sun, 30 May 2004 13:09:21 +0000 (06:09 -0700)]
[PATCH] sparse: n_tty annotation

n_tty annotated, ldisc_struct ->read() and ->write() got __user
on their buf argument.

22 years ago[PATCH] sparse: vt and friends
Alexander Viro [Sun, 30 May 2004 13:09:10 +0000 (06:09 -0700)]
[PATCH] sparse: vt and friends

vt, vt_ioctl, consolemap and selection annotated, struct unimapdesc
and struct consolefontdesc got __user on their pointer members.

22 years ago[PATCH] sparse: tty_ioctl annotation
Alexander Viro [Sun, 30 May 2004 13:08:59 +0000 (06:08 -0700)]
[PATCH] sparse: tty_ioctl annotation

tty_ioctl.c annotation

22 years ago[PATCH] sparse: tty_io annotation
Alexander Viro [Sun, 30 May 2004 13:08:48 +0000 (06:08 -0700)]
[PATCH] sparse: tty_io annotation

tty_io.c annotated

22 years ago[PATCH] sparse: fs/compat.c::copy_iocb() cleanup
Alexander Viro [Sun, 30 May 2004 13:08:37 +0000 (06:08 -0700)]
[PATCH] sparse: fs/compat.c::copy_iocb() cleanup

That "u64" in there is actually struct iocb __user *.  Changed
prototype, killed casts in body and caller...

22 years ago[PATCH] sparse: amd64 - rest
Alexander Viro [Sun, 30 May 2004 13:08:26 +0000 (06:08 -0700)]
[PATCH] sparse: amd64 - rest

return void in floppy.h and "conditionally long" constants in page.h.

Would be useful to teach as(1) about C notation for long...

22 years ago[PATCH] sparse: amd64 - long constants
Alexander Viro [Sun, 30 May 2004 13:08:15 +0000 (06:08 -0700)]
[PATCH] sparse: amd64 - long constants

Long constants marked as such.

22 years ago[PATCH] sparse: amd64 - #if abuse fixes
Alexander Viro [Sun, 30 May 2004 13:08:04 +0000 (06:08 -0700)]
[PATCH] sparse: amd64 - #if abuse fixes

#if where #ifdef should've been.  Fixed.

22 years ago[PATCH] sparse: amd64 - more trivial annotations
Alexander Viro [Sun, 30 May 2004 13:07:53 +0000 (06:07 -0700)]
[PATCH] sparse: amd64 - more trivial annotations

22 years ago[PATCH] sparse: amd64 annotations - trivial part
Alexander Viro [Sun, 30 May 2004 13:07:42 +0000 (06:07 -0700)]
[PATCH] sparse: amd64 annotations - trivial part

Annotation of amd64 code - just adding __user where needed...

22 years ago[PATCH] sparse: amd64 annotation - beginning
Alexander Viro [Sun, 30 May 2004 13:07:30 +0000 (06:07 -0700)]
[PATCH] sparse: amd64 annotation - beginning

Annotated basic primitives (copy_.._user, get_user, ...).
Functions got __user in prototypes, macros (in uaccess.h) got a
conditional check - see CHECK_UPTR() below.  If you have more elegant
way to deal with the macros - please, tell.

22 years ago[PATCH] sparse: net/* non-ANSI argument lists
Alexander Viro [Sun, 30 May 2004 13:07:19 +0000 (06:07 -0700)]
[PATCH] sparse: net/* non-ANSI argument lists

22 years ago[PATCH] sparse: more net/* annotation
Alexander Viro [Sun, 30 May 2004 13:07:08 +0000 (06:07 -0700)]
[PATCH] sparse: more net/* annotation

Annotation of net/* ...copy... that deals with userland pointers +
several bits that fell out of previous patch (trivial annotations)

22 years ago[PATCH] sparse: trivial part of net/* annotation
Alexander Viro [Sun, 30 May 2004 13:06:56 +0000 (06:06 -0700)]
[PATCH] sparse: trivial part of net/* annotation

22 years ago[PATCH] sparse: missed piece of sock_get_timestamp() annotation
Alexander Viro [Sun, 30 May 2004 13:06:44 +0000 (06:06 -0700)]
[PATCH] sparse: missed piece of sock_get_timestamp() annotation

22 years ago[PATCH] sparse: dev_ioctl() callers
Alexander Viro [Sun, 30 May 2004 13:06:33 +0000 (06:06 -0700)]
[PATCH] sparse: dev_ioctl() callers

dev_ioctl() has a __user pointer argument; however, its declaration
lacks that and callers are also wrong.  Declaration fixed, callers updated.

22 years ago[PATCH] sparse: missed setsockopt wrappers
Alexander Viro [Sun, 30 May 2004 13:06:21 +0000 (06:06 -0700)]
[PATCH] sparse: missed setsockopt wrappers

Here's the tail of setsockopt I've missed - for old protocol families we
have wrappers around those methods.

22 years ago[PATCH] Fix compilation on x86-64
Andi Kleen [Sun, 30 May 2004 13:01:28 +0000 (06:01 -0700)]
[PATCH] Fix compilation on x86-64

I managed to break compilation with my last patch with a last minute edit.

22 years agoMerge davem@nuts.davemloft.net:/disk1/BK/net-2.6
David S. Miller [Sun, 30 May 2004 12:22:41 +0000 (05:22 -0700)]
Merge davem@nuts.davemloft.net:/disk1/BK/net-2.6
into kernel.bkbits.net:/home/davem/net-2.6

22 years ago[PKT_SCHED]: Missing rta_len init in sch_delay.
David S. Miller [Sun, 30 May 2004 12:12:54 +0000 (05:12 -0700)]
[PKT_SCHED]: Missing rta_len init in sch_delay.

22 years agoMerge davem@nuts.davemloft.net:/disk1/BK/sparc-2.6
David S. Miller [Sun, 30 May 2004 10:51:09 +0000 (03:51 -0700)]
Merge davem@nuts.davemloft.net:/disk1/BK/sparc-2.6
into kernel.bkbits.net:/home/davem/sparc-2.6

22 years ago[SPARC]: Do tty_flip_buffer_push outside of port lock.
David S. Miller [Sun, 30 May 2004 10:43:22 +0000 (03:43 -0700)]
[SPARC]: Do tty_flip_buffer_push outside of port lock.

22 years ago[SPARC64]: Update defconfig.
David S. Miller [Sun, 30 May 2004 10:39:32 +0000 (03:39 -0700)]
[SPARC64]: Update defconfig.

22 years agoppc64: use "ASM_CONST()" to give proper C type to constants
Linus Torvalds [Sun, 30 May 2004 05:50:19 +0000 (22:50 -0700)]
ppc64: use "ASM_CONST()" to give proper C type to constants
that can also be used in assembly language context.

This one fixes the common CPU_FTR_COHERENT_ICACHE warning in
asm/cacheflush.h.

22 years ago[PATCH] Really fix empty node 0 on x86-64
Andi Kleen [Sun, 30 May 2004 04:55:25 +0000 (21:55 -0700)]
[PATCH] Really fix empty node 0 on x86-64

x86-64 shares some numa code with i386, and that didn't handle holes in
the node map properly.  Fix this.

22 years ago[PATCH] x86-64 update
Andi Kleen [Sun, 30 May 2004 04:55:13 +0000 (21:55 -0700)]
[PATCH] x86-64 update

Add accumulated bug fixes for x86-64 and some minor cleanups.

 - Add ptep_set_access_flags from i386
 - Change memory text mapping to 10MB from 40MB to make alias window
   smaller
 - Disable change_page_attr for kernel mapoing for now (wrong, but the
   other code also doesn't work)
 - Add .bss.page_aligned and align interrupt/exception stacks (idea from
   Matt Mackall)
 - Protect sysenter MSR setup against missing registers.
 - Set mce tolerance level to 1 (panic instead of deadlock)
 - Fix cross 4GB bug in find_first_bit (Michael Matz)
 - Add missing memory clobbers in bitops.h (Suresh B.  Siddha)
 - Check kernel size at boot up
 - Set boot cpu online at boot to fix early printk
 - Fix EFER bit definitions (Dave Jones)
 - Fix comments in 32bit syscall table.
 - Quieten all 32bit syscalls that are unimplemented in 32bit.
 - Various MCE fixes suggested by Eric Morton, David Boles et.al.
   More to come.
 - Quieten some printks
 - Fix cpu online check in msr driver
 - Lower polling interval for machine checks to 5 minutes to avoid
   overrunning the registers.
 - Make memcpy_{from,to}io more compatible with i386 (Roland Dreier)

22 years agoppc64: fix untyped large constants so that they don't cause
Linus Torvalds [Sun, 30 May 2004 04:47:20 +0000 (21:47 -0700)]
ppc64: fix untyped large constants so that they don't cause
sparse to warn about implicit typing.

22 years agoAdd __user annotations to sock_get_timestamp()
Linus Torvalds [Sun, 30 May 2004 04:10:51 +0000 (21:10 -0700)]
Add __user annotations to sock_get_timestamp()

22 years ago[PATCH] make vram boot option actually work
Piotr Kaczuba [Sun, 30 May 2004 03:32:16 +0000 (20:32 -0700)]
[PATCH] make vram boot option actually work

The following patch makes the vram boot option actually be recognized
and its value assigned to the vram variable.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
22 years agoLinux 2.6.7-rc2 v2.6.7-rc2
Linus Torvalds [Sat, 29 May 2004 16:24:11 +0000 (09:24 -0700)]
Linux 2.6.7-rc2

22 years ago[PATCH] sparse: nf_sockopt() annotation
Alexander Viro [Sat, 29 May 2004 12:36:13 +0000 (05:36 -0700)]
[PATCH] sparse: nf_sockopt() annotation

netfilter set/getsockopt annotated.  That's a continuation of previous
patch - by now the only remaining place where setsockopt()/getsockopt()
userland pointers are not traced is vcc_[sg]etsockopt() and down into the
ATM drivers.  Everything from that area in net/* should be done by now.

That's pretty much it as far as non-trivial splitting counts; remaining
patches in net/* are independent, tty and console stuff is also not hard to
split (not much of it, anyway) and what remains is arch-dependent code for
alpha and amd64.  Of course there'll be more stuff coming, but as far as I'm
concerned the rest of my pile is managable - and finally below 200Kb.  That'll
go tomorrow - right now I'm going down and getting some sleep...

22 years agoMerge Al/Russell fighting over the same __user annotations
Linus Torvalds [Sat, 29 May 2004 12:33:21 +0000 (05:33 -0700)]
Merge Al/Russell fighting over the same __user annotations

22 years ago[SERIAL] Fix sparse warnings in serial_core.c
Russell King [Sat, 29 May 2004 21:24:42 +0000 (22:24 +0100)]
[SERIAL] Fix sparse warnings in serial_core.c

This fixes sparse warnings for the user write method, and the UART
ioctl functions which copy data to/from userspace.

22 years agoMerge bk://bk.arm.linux.org.uk/linux-2.6-rmk
Linus Torvalds [Sat, 29 May 2004 12:27:15 +0000 (05:27 -0700)]
Merge bk://bk.arm.linux.org.uk/linux-2.6-rmk
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years ago[ARM PATCH] 1897/1: prevent selecting more than one PXA target
Nicolas Pitre [Sat, 29 May 2004 21:54:21 +0000 (22:54 +0100)]
[ARM PATCH] 1897/1: prevent selecting more than one PXA target

Patch from Nicolas Pitre

... since this is not supported at the moment and would fail to build.

22 years ago[ARM PATCH] 1872/1: base clock difference between PXA25x and PXA27x
Nicolas Pitre [Sat, 29 May 2004 21:50:19 +0000 (22:50 +0100)]
[ARM PATCH] 1872/1: base clock difference between PXA25x and PXA27x

Patch from Nicolas Pitre

22 years ago[ARM PATCH] 1898/1: fix io_v2p macro on PXA
Nicolas Pitre [Sat, 29 May 2004 21:41:43 +0000 (22:41 +0100)]
[ARM PATCH] 1898/1: fix io_v2p macro on PXA

Patch from Nicolas Pitre

Correcting my own screw-up, found by Ian Campbell.

22 years ago[ARM] Remove obsolete asm-arm/arch-*/keyboard.h
Russell King [Sat, 29 May 2004 12:44:29 +0000 (13:44 +0100)]
[ARM] Remove obsolete asm-arm/arch-*/keyboard.h

22 years agoMerge bk://bk.arm.linux.org.uk/linux-2.6-pcmcia
Linus Torvalds [Sat, 29 May 2004 12:25:38 +0000 (05:25 -0700)]
Merge bk://bk.arm.linux.org.uk/linux-2.6-pcmcia
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years ago[ARM PATCH] 1869/1: PCMCIA support for Mainstone
Nicolas Pitre [Sat, 29 May 2004 22:03:57 +0000 (23:03 +0100)]
[ARM PATCH] 1869/1: PCMCIA support for Mainstone

Patch from Nicolas Pitre

22 years ago[ARM] Ensure AMBA devices on Versatile have a correct dma_mask
Russell King [Sat, 29 May 2004 12:12:55 +0000 (13:12 +0100)]
[ARM] Ensure AMBA devices on Versatile have a correct dma_mask

22 years ago[PATCH] sparse: ->[gs]etsockopt() annotation
Alexander Viro [Sat, 29 May 2004 12:02:23 +0000 (05:02 -0700)]
[PATCH] sparse: ->[gs]etsockopt() annotation

optval (and in case of getsockopt - optlen) made __user, changes
percolated down into the instances.

22 years agoFix up my misedit of Al's good patch. Don't blame Al.
Linus Torvalds [Sat, 29 May 2004 12:00:35 +0000 (05:00 -0700)]
Fix up my misedit of Al's good patch. Don't blame Al.

22 years ago[ARM] Move common definitions to asm/memory.h
Russell King [Sat, 29 May 2004 11:53:46 +0000 (12:53 +0100)]
[ARM] Move common definitions to asm/memory.h

Virtually all the architecture specific memory.h files were defining
the same constants and v<->p translations the same way.  Move them to
asm/memory.h

22 years ago[PATCH] sparse: final bits of compat_ioctl
Alexander Viro [Sat, 29 May 2004 11:44:03 +0000 (04:44 -0700)]
[PATCH] sparse: final bits of compat_ioctl

ethtool_ioctl() annotated; it would go in the first compat_ioctl
chunk (trivial ones) if it wouldn't be right after SIOCGIFCONF mess; was
easier to split in that order...

22 years ago[PATCH] sparse (compat_ioctl): SIOCGIFCONF
Alexander Viro [Sat, 29 May 2004 11:43:52 +0000 (04:43 -0700)]
[PATCH] sparse (compat_ioctl): SIOCGIFCONF

Switched to compat_alloc_user_space(), killed kmalloc and set_fs(),
annotated, somewhat cleaned up.

22 years ago[PATCH] sparse (compat_ioctl): fb colourmap stuff
Alexander Viro [Sat, 29 May 2004 11:43:41 +0000 (04:43 -0700)]
[PATCH] sparse (compat_ioctl): fb colourmap stuff

Similar to previous - annotated, killed helper, made the thing
type-clean, cleaned up.

22 years ago[PATCH] sparse (compat_ioctl): CDROM_SEND_PACKET handling
Alexander Viro [Sat, 29 May 2004 11:43:31 +0000 (04:43 -0700)]
[PATCH] sparse (compat_ioctl): CDROM_SEND_PACKET handling

annotated, killed useless (and type-buggering - void ** is not
assignment-compatible with foo **) helper, cleaned up.

22 years ago[PATCH] sparse (compat_ioctl): font stuff
Alexander Viro [Sat, 29 May 2004 11:43:20 +0000 (04:43 -0700)]
[PATCH] sparse (compat_ioctl): font stuff

Annotated, cleaned up, wholesale copying of structure with
subsequeunt casts from hell replaced with get_user()/put_user() on
the fields we really need.

22 years ago[PATCH] sparse (compat_ioctl): blkpg cleanup and annotation
Alexander Viro [Sat, 29 May 2004 11:43:09 +0000 (04:43 -0700)]
[PATCH] sparse (compat_ioctl): blkpg cleanup and annotation

blkpg ioctls handling switched to compat_alloc_user_space() (instead
of messing with set_fs()) and annotated.

22 years ago[PATCH] sparse (compat_ioctl): raw_ioctl() fixes
Alexander Viro [Sat, 29 May 2004 11:42:59 +0000 (04:42 -0700)]
[PATCH] sparse (compat_ioctl): raw_ioctl() fixes

Whoever had done the translation apparently never heard of
big-endian boxen.  Replaced the kludge with get_user()/put_user() on
64bit values.  Note that
struct raw_config_request
{
        int     raw_minor;
        __u64   block_major;
        __u64   block_minor;
};
has __u64 on _all_ platforms and it is used as an integer in raw.c code.
The only reason why translation is needed at all is alignment for 64bit
values; block_major and block_minor themselves should be passed as-is.

Breakage fixed, entire area annotated.

22 years ago[PATCH] sparse: compat_ioctl - easy part
Alexander Viro [Sat, 29 May 2004 11:42:48 +0000 (04:42 -0700)]
[PATCH] sparse: compat_ioctl - easy part

Easy parts of compat_ioctl fixes - pure __user annotation.
In one place (SG_IO handling) replaced dereference of user pointer
with put_user() (why whoever had noticed that just put a FIXME instead
of obvios fix is a mistery...)

22 years ago[PATCH] sparse: fault_in_pages_readable() annotation
Alexander Viro [Sat, 29 May 2004 09:18:33 +0000 (02:18 -0700)]
[PATCH] sparse: fault_in_pages_readable() annotation

22 years ago[PATCH] sparse: misc ->read()/->write() __user annotation
Alexander Viro [Sat, 29 May 2004 09:18:23 +0000 (02:18 -0700)]
[PATCH] sparse: misc ->read()/->write() __user annotation

a bunch of trivial ->read() and ->write() instances (and their procfs
relatives) __user-annotated.  All chunks are independent, but IMO not
worth splitting.

22 years ago[PATCH] sparse: bits and pieces
Alexander Viro [Sat, 29 May 2004 09:18:12 +0000 (02:18 -0700)]
[PATCH] sparse: bits and pieces

Independent minor bits caught by sparse:

 - paride.h mixing void and int in ? :, used always in a void context
   ide-iops.c return insw() - insw is void()
 - scsi/constants.c uses undefined macros in #if; added #define to 0 in
   case that used to leave it undefined
 - usb/host/hcd.h: fixed-point arithmetics in constant
 - fs/exec.c: missing UL on a large constant
 - fs/locks.c: #if where #ifdef should've been
 - fs.h: missing UL on MAX_LFS_FILESIZE in 64bit case

22 years ago[PATCH] sparse: trivial parts of fs/* annotation
Alexander Viro [Sat, 29 May 2004 09:18:01 +0000 (02:18 -0700)]
[PATCH] sparse: trivial parts of fs/* annotation

fs/*, fs/proc/*, fs/nfsd/*, fs/autofs4/*

The only interdependent group is the fs/autofs4/* stuff.

22 years ago[PATCH] sparse: aio annotation
Alexander Viro [Sat, 29 May 2004 09:17:49 +0000 (02:17 -0700)]
[PATCH] sparse: aio annotation

22 years ago[PATCH] sparse: sg.c annotation
Alexander Viro [Sat, 29 May 2004 09:17:39 +0000 (02:17 -0700)]
[PATCH] sparse: sg.c annotation

rest of pointers in sg ioctl arguments made __user.

22 years ago[PATCH] sparse: scsi_ioctl __user annotation
Alexander Viro [Sat, 29 May 2004 09:17:28 +0000 (02:17 -0700)]
[PATCH] sparse: scsi_ioctl __user annotation

scsi_ioctl() and scsi_ioctl_send_command() switched to __user *;
scsi/scsi_ioctl.c annotated.

22 years ago[PATCH] sparse: scsi_cmd_ioctl __user annotation
Alexander Viro [Sat, 29 May 2004 09:17:16 +0000 (02:17 -0700)]
[PATCH] sparse: scsi_cmd_ioctl __user annotation

scsi_cmd_ioctl() switched to __user *, block/scsi_ioctl.c annotated.

22 years ago[PATCH] sparse: make sg_io_hdr->cmdp a __user pointer
Alexander Viro [Sat, 29 May 2004 09:17:05 +0000 (02:17 -0700)]
[PATCH] sparse: make sg_io_hdr->cmdp a __user pointer

copying moved into sg_io()

22 years ago[PATCH] sparse: kernel/sysctl.c annotation and cleanup
Alexander Viro [Sat, 29 May 2004 09:16:55 +0000 (02:16 -0700)]
[PATCH] sparse: kernel/sysctl.c annotation and cleanup

22 years ago[PATCH] sparse: trivial part of cdrom.c annotation
Alexander Viro [Sat, 29 May 2004 09:16:44 +0000 (02:16 -0700)]
[PATCH] sparse: trivial part of cdrom.c annotation

22 years ago[PATCH] sparse: trivial part of drivers/scsi/* annotation
Alexander Viro [Sat, 29 May 2004 09:16:34 +0000 (02:16 -0700)]
[PATCH] sparse: trivial part of drivers/scsi/* annotation

22 years ago[PATCH] sparse: cdrom_generic_command split into kernel/user
Alexander Viro [Sat, 29 May 2004 09:16:22 +0000 (02:16 -0700)]
[PATCH] sparse: cdrom_generic_command split into kernel/user

The kernel instance (packet_command) has kernel pointers, userland one -
__user ones. Surprise.

22 years ago[PATCH] more sparse checking of do_utimes()
Alexander Viro [Sat, 29 May 2004 07:33:30 +0000 (00:33 -0700)]
[PATCH] more sparse checking of do_utimes()

Speaking of conflicts...  Here's another half of a do_utimes() patch,
same story as on ppc64:

22 years agoAdd __user pointer annotations to ppc64 code
Linus Torvalds [Sat, 29 May 2004 07:32:11 +0000 (00:32 -0700)]
Add __user pointer annotations to ppc64 code

22 years agoMerge bk://cifs.bkbits.net/linux-2.5cifs
Linus Torvalds [Sat, 29 May 2004 06:39:07 +0000 (23:39 -0700)]
Merge bk://cifs.bkbits.net/linux-2.5cifs
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years agoMerge bk://bk.linux1394.org/ieee1394-2.6
Linus Torvalds [Sat, 29 May 2004 06:37:49 +0000 (23:37 -0700)]
Merge bk://bk.linux1394.org/ieee1394-2.6
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years agoadb.c needs <linux/devfs_fs_kernel.h>
Linus Torvalds [Sat, 29 May 2004 06:33:01 +0000 (23:33 -0700)]
adb.c needs <linux/devfs_fs_kernel.h>

22 years agoMerge bk://kernel.bkbits.net/davem/net-2.6
Linus Torvalds [Sat, 29 May 2004 06:26:05 +0000 (23:26 -0700)]
Merge bk://kernel.bkbits.net/davem/net-2.6
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years agoMerge bk://kernel.bkbits.net/davem/sparc-2.6
Linus Torvalds [Sat, 29 May 2004 06:25:16 +0000 (23:25 -0700)]
Merge bk://kernel.bkbits.net/davem/sparc-2.6
into ppc970.osdl.org:/home/torvalds/v2.6/linux

22 years ago[PATCH] Missed RTC_IRQ patchlet
Alexander Viro [Sat, 29 May 2004 06:18:42 +0000 (23:18 -0700)]
[PATCH] Missed RTC_IRQ patchlet

oh, hell - I missed this one when carving the thing out of big patch ;-/