[PATCH 0/4] m32r: build fixes

[PATCH 0/4] m32r: build fixes

[Kyle McMartin]

Hi Linus,

Not sure if I should send this directly to you or not, but it seems like
m32r has not received much love recently. The following patches fix the
defconfig build of m32r for me (which I needed to do to test the
__do_IRQ removal.)

regards, Kyle

[PATCH 1/4] m32r: restore _BLOCKABLE

[Kyle McMartin]

Commit a7f8388e accidentally removed it...

Signed-off-by: Kyle McMartin <kyle@redhat.com>
---
 arch/m32r/kernel/signal.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/arch/m32r/kernel/signal.c b/arch/m32r/kernel/signal.c
index 7bbe386..cc36fe1 100644
--- a/arch/m32r/kernel/signal.c
+++ b/arch/m32r/kernel/signal.c
@@ -28,6 +28,8 @@
 
 #define DEBUG_SIG 0
 
+#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
+
 asmlinkage int
 sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
 		unsigned long r2, unsigned long r3, unsigned long r4,
-- 
1.7.3.1

[PATCH 2/4] m32r: get_user takes an lvalue, not a pointer

[Kyle McMartin]

Signed-off-by: Kyle McMartin <kyle@redhat.com>
---
 arch/m32r/kernel/signal.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/m32r/kernel/signal.c b/arch/m32r/kernel/signal.c
index cc36fe1..a08697f 100644
--- a/arch/m32r/kernel/signal.c
+++ b/arch/m32r/kernel/signal.c
@@ -256,7 +256,7 @@ give_sigsegv:
 static int prev_insn(struct pt_regs *regs)
 {
 	u16 inst;
-	if (get_user(&inst, (u16 __user *)(regs->bpc - 2)))
+	if (get_user(inst, (u16 __user *)(regs->bpc - 2)))
 		return -EFAULT;
 	if ((inst & 0xfff0) == 0x10f0)	/* trap ? */
 		regs->bpc -= 2;
-- 
1.7.3.1

[PATCH 3/4] m32r: add kernel/.gitignore and ignore vmlinux.lds

[Kyle McMartin]

Signed-off-by: Kyle McMartin <kyle@redhat.com>
---
 arch/m32r/kernel/.gitignore |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)
 create mode 100644 arch/m32r/kernel/.gitignore

diff --git a/arch/m32r/kernel/.gitignore b/arch/m32r/kernel/.gitignore
new file mode 100644
index 0000000..c5f676c
--- /dev/null
+++ b/arch/m32r/kernel/.gitignore
@@ -0,0 +1 @@
+vmlinux.lds
-- 
1.7.3.1

[PATCH 4/4] m32r: test __LITTLE_ENDIAN__ instead of __LITTLE_ENDIAN

[Kyle McMartin]

Fixes build for me... these are what's tested in byteorder.h...

Signed-off-by: Kyle McMartin <kyle@redhat.com>
---
 arch/m32r/include/asm/elf.h |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/m32r/include/asm/elf.h b/arch/m32r/include/asm/elf.h
index 2f85412..b8da7d0 100644
--- a/arch/m32r/include/asm/elf.h
+++ b/arch/m32r/include/asm/elf.h
@@ -82,9 +82,9 @@ typedef elf_fpreg_t elf_fpregset_t;
  * These are used to set parameters in the core dumps.
  */
 #define ELF_CLASS	ELFCLASS32
-#if defined(__LITTLE_ENDIAN)
+#if defined(__LITTLE_ENDIAN__)
 #define ELF_DATA	ELFDATA2LSB
-#elif defined(__BIG_ENDIAN)
+#elif defined(__BIG_ENDIAN__)
 #define ELF_DATA	ELFDATA2MSB
 #else
 #error no endian defined
-- 
1.7.3.1

Re: [PATCH 1/4] m32r: restore _BLOCKABLE

[Al Viro]

On Fri, Oct 15, 2010 at 09:16:45PM -0400, Kyle McMartin wrote:
> Commit a7f8388e accidentally removed it...
> 
> Signed-off-by: Kyle McMartin <kyle@redhat.com>
> ---
>  arch/m32r/kernel/signal.c |    2 ++
>  1 files changed, 2 insertions(+), 0 deletions(-)
> 
> diff --git a/arch/m32r/kernel/signal.c b/arch/m32r/kernel/signal.c
> index 7bbe386..cc36fe1 100644
> --- a/arch/m32r/kernel/signal.c
> +++ b/arch/m32r/kernel/signal.c
> @@ -28,6 +28,8 @@
>  
>  #define DEBUG_SIG 0
>  
> +#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
> +

Grrr...   Sorry, reordering breakage.  In the signals tree here I have

static inline void sig_set_blocked(struct sigset_t *set)
{
        sigdelsetmask(set, sigmask(SIGKILL) | sigmask(SIGSTOP));
        spin_lock_irq(&current->sighand->siglock);
        current->blocked = *set;
        recalc_sigpending();
        spin_unlock_irq(&current->sighand->siglock);
}

and it's used all over the place (including quite a few places where
we currently have sigprocmask(SIG_SETMASK, set, NULL), which is what
it's equivalent to).  With that done, m32r doesn't use _BLOCKABLE
anywhere, so it got removed.  And that chunk got picked when I'd been
reordering the queue to pull the arch-specific fixes in front.  Sorry.

Re: [PATCH 2/4] m32r: get_user takes an lvalue, not a pointer

[Al Viro]

On Fri, Oct 15, 2010 at 09:17:09PM -0400, Kyle McMartin wrote:
>  static int prev_insn(struct pt_regs *regs)
>  {
>  	u16 inst;
> -	if (get_user(&inst, (u16 __user *)(regs->bpc - 2)))
> +	if (get_user(inst, (u16 __user *)(regs->bpc - 2)))

Ouch...  ACKed, my fuckup.