linuxsh-dev

Re: PATA Patch for Hp6xx

[Paul M. <le...@li...>]

On Fri, Mar 02, 2007 at 07:06:09AM +0100, Manuel Lauss wrote:
> On Thu, Mar 01, 2007 at 11:43:28PM +0000, Kristoffer Ericson wrote:
> > +	              .start = 0x1f0 + 0x206,               <------------ Error?
> > +		      .end   = 0x1f0 +8 + 0x206 + 8,
> > +		      .flags = IORESOURCE_IO,
> 
> 3f6 is correct. .end should be 3f7. (have alook at
> arch/sh/cchips/hd6446x/hd64461/io.c::PORT2ADDR()).
> 
> Maybe you could even use the "real" MMIO addresses of
> the hd64461 pcmcia0 region (0xb50001f0), and one day get rid of
> that ugly io.c file altogether...
> 
Indeed, and in this case, there's really no reason _not_ to use the MMIO
address. The reason some folks were using the PIO addresses before were
as a lame way to take care of port shifting, and this is really the wrong
way to go about it. If there's a real inability to do an access of that
size, then it's something the driver has to deal with, we shouldn't be
papering over it in the machvec.

I would say rip out the HD64461 entries from the machvec entirely and see
what falls out. I imagine PCMCIA will be a problem, but this needs an
overhaul for the hardirq changes anyways, and then we're likely left with
something much smaller that we can push in as an MFD driver, and kill off
the rest of the cchip mess.

The io.c's have always been utter crap, the sooner we can kill off the
outstanding ones, the better.

Re: PATA Patch for Hp6xx

[Manuel L. <ma...@ro...>]

On Thu, Mar 01, 2007 at 11:43:28PM +0000, Kristoffer Ericson wrote:
> +	              .start = 0x1f0 + 0x206,               <------------ Error?
> +		      .end   = 0x1f0 +8 + 0x206 + 8,
> +		      .flags = IORESOURCE_IO,

3f6 is correct. .end should be 3f7. (have alook at
arch/sh/cchips/hd6446x/hd64461/io.c::PORT2ADDR()).

Maybe you could even use the "real" MMIO addresses of
the hd64461 pcmcia0 region (0xb50001f0), and one day get rid of
that ugly io.c file altogether...

-- 
 ml.

PATA Patch for Hp6xx

[Kristoffer E. <kri...@ho...>]

(forgot to attach patch)

#include <asm/irq.h>
#include <asm/hp6xx.h>
#include <asm/cpu/dac.h>
+#include <linux/platform_device.h>

#define	SCPCR	0xa4000116
#define SCPDR	0xa4000136

+static struct resource cf_ide_resources[] = {
+              [0] = {
+	              .start = 0x1f0,
+		      .end   = 0x1f0 + 8,
+		      .flags = IORESOURCE_IO,
+		      },
+	      [1] = {
+	              .start = 0x1f0 + 0x206,               <------------ Error?
+		      .end   = 0x1f0 +8 + 0x206 + 8,
+		      .flags = IORESOURCE_IO,
+		    },
+              [2] = {
+	              .start = 93,
+		      .flags = IORESOURCE_IRQ,
+		    },
+};
+
+static struct platform_device cf_ide_device = {
+              .name         =  "pata_platform",
+	      .id           =  -1,
+	      .num_resources = ARRAY_SIZE(cf_ide_resources),
+	      .resource      = cf_ide_resources,
+};
+
+static struct platform_device *hp6xx_devices[] __initdata = {
+
+       &cf_ide_device,
+};
+
+static int __init hp6xx_devices_setup(void)
+{
+ return platform_add_devices(hp6xx_devices,ARRAY_SIZE(hp6xx_devices));
+}
+
static void __init hp6xx_setup(char **cmdline_p)
{
	u8 v8;
	u16 v;

+        device_initcall(hp6xx_devices_setup);
+
	v = inw(HD64461_STBCR);
	v |= HD64461_STBCR_SURTST | HD64461_STBCR_SIRST |
	    HD64461_STBCR_STM1ST | HD64461_STBCR_STM0ST |

_________________________________________________________________
Don't just search. Find. Check out the new MSN Search! 
http://search.msn.com/

HP6xx 2.6.20 kernel producing unaligned access errors

[Kristoffer E. <kri...@ho...>]

Greetings,

I tried to get pata support working properly and it now detects the CF slot 
and the sandisk card inside it (with correct size). Upon bootup (from root 
device) it produces these errors:

Freeing unused kernel memory: 68k freed
Fixing up unaligned userspace access in "hotplug" pid=181 pc=0x000141ea 
ins=0x5436
Fixing up unaligned userspace access in "hotplug" pid=181 pc=0x000141ee 
ins=0x0203
Sending SIGBUS to "hotplug" due to unaligned access (PC 141f1 PR 141f2)
Fixing up unaligned userspace access in "hotplug" pid=182 pc=0x000141ea 
ins=0x5436
Fixing up unaligned userspace access in "hotplug" pid=182 pc=0x000141ee 
ins=0x0203
Sending SIGBUS to "hotplug" due to unaligned access (PC 141f1 PR 141f2)

I've attached the pata patch and please note that the only driver left to 
implement is the hp6xx keyboard (if that has anything to do with it for 
whatever reason).

Best wishes
Kristoffer Ericson

_________________________________________________________________
Don't just search. Find. Check out the new MSN Search! 
http://search.msn.click-url.com/go/onm00200636ave/direct/01/

Re: duplicated syscall checks

[Paul M. <le...@li...>]

On Thu, Mar 01, 2007 at 11:52:41AM +0900, Hideo Saito wrote:
> When '-ERESTARTNOINTR' is returned from fork syscall, do_signal()
> subtracts 2 from the address that is handled by alarm signal.
> 
This was already spotted and fixed in current git by Sakato-san, in the
case where we actually have a signal we should simply be returning.

I'll send this off for inclusion in the -stable tree.

commit fade884146b44be85b0987c8357194a16ea72f90
Author: Ryusuke Sakato <sa...@hs...>
Date:   Fri Feb 23 13:22:56 2007 +0900

    sh: Fix sigmask trampling in signal delivery.
    
    There was a missing return in do_signal() that caused the saved
    sigmask to be written back after having successfully delivered
    the signal.
    
    Signed-off-by: Ryusuke Sakato <sa...@hs...>
    Signed-off-by: Paul Mundt <le...@li...>

diff --git a/arch/sh/kernel/signal.c b/arch/sh/kernel/signal.c
index 32f10a0..9f39ef1 100644
--- a/arch/sh/kernel/signal.c
+++ b/arch/sh/kernel/signal.c
@@ -589,6 +589,8 @@ static void do_signal(struct pt_regs *regs, unsigned int save_r0)
 			if (test_thread_flag(TIF_RESTORE_SIGMASK))
 				clear_thread_flag(TIF_RESTORE_SIGMASK);
 		}
+
+		return;
 	}
 
  no_signal:
@@ -598,7 +600,7 @@ static void do_signal(struct pt_regs *regs, unsigned int save_r0)
 		if (regs->regs[0] == -ERESTARTNOHAND ||
 		    regs->regs[0] == -ERESTARTSYS ||
 		    regs->regs[0] == -ERESTARTNOINTR) {
-		    	regs->regs[0] = save_r0;
+			regs->regs[0] = save_r0;
 			regs->pc -= 2;
 		} else if (regs->regs[0] == -ERESTART_RESTARTBLOCK) {
 			regs->pc -= 2;

duplicated syscall checks

[Hideo S. <sa...@de...>]

Hi Paul,

When I run spawn that is one of benchmark programs on unixbench-4.1.0, an error is asserted as follows.

% spawn 30
Illegal instruction

The caused PC is 0x4007fe, but the PC should be 0x400800, please refer to appended spawn's disassembled list. The 'report' is handled by alarm signal.

When '-ERESTARTNOINTR' is returned from fork syscall, do_signal() subtracts 2 from the address that is handled by alarm signal.

I think that the following patch can avoid this problem because the handled function does not come from a system call.

--- arch/sh/kernel/signal.c.org	Mon Feb  5 03:44:54 2007
+++ arch/sh/kernel/signal.c	Thu Mar  1 11:14:54 2007
@@ -583,16 +583,17 @@ static void do_signal(struct pt_regs *re
 		/* Whee!  Actually deliver the signal.  */
 		if (handle_signal(signr, &ka, &info, oldset, regs) == 0) {
 			/* a signal was successfully delivered; the saved
 			 * sigmask will have been stored in the signal frame,
 			 * and will be restored by sigreturn, so we can simply
 			 * clear the TIF_RESTORE_SIGMASK flag */
 			if (test_thread_flag(TIF_RESTORE_SIGMASK))
 				clear_thread_flag(TIF_RESTORE_SIGMASK);
+			regs->tra = -1;		/* disable syscall checks */
 		}
 	}
 
  no_signal:
 	/* Did we come from a system call? */
 	if (regs->tra >= 0) {
 		/* Restart the system call - no handlers present */
 		if (regs->regs[0] == -ERESTARTNOHAND ||

-a part of spawn's disassembled list----------------------------------
wake_me:
          004007e0 d005  MOV.L   @(H'14,PC),R0   ; 0x4007f8;*__sysv_signal@@GLIB
C_2.2
          004007e2 2f86  MOV.L   R8,@-R15        
          004007e4 6843  MOV     R4,R8           
          004007e6 4f22  STS.L   PR,@-R15        
          004007e8 400b  JSR     @R0             
          004007ea e40e  MOV     #H'E,R4         
          004007ec 4f26  LDS.L   @R15+,PR        
          004007ee d103  MOV.L   @(H'C,PC),R1    ; 0x4007fc;*alarm@@GLIBC_2.2
          004007f0 6483  MOV     R8,R4           
          004007f2 412b  JMP     @R1             
          004007f4 68f6  MOV.L   @R15+,R8        
          004007f6 0009  NOP                     
          004007f8 057c  MOV.B   @(R0,R7),R5     ; __sysv_signal@@GLIBC_2.2
          004007fa 0040  .data.w h'0040          
          004007fc 0608  .data.w h'0608          ; alarm@@GLIBC_2.2
          004007fe 0040  .data.w h'0040          
report:
          00400800 4f22  STS.L   PR,@-R15        
          00400802 d105  MOV.L   @(H'14,PC),R1   ; 0x400818;*_edata
          00400804 d005  MOV.L   @(H'14,PC),R0   ; 0x40081c;*fprintf@@GLIBC_2.2

Re: pdflush can't exit

[Paul M. <le...@li...>]

On Tue, Feb 27, 2007 at 11:27:04AM +0900, Hideo Saito wrote:
> On Mon, 26 Feb 2007 18:38:46 +0900, Paul Mundt wrote:
> > Yes, I think this is the right fix. Can you please provide a
> > Signed-off-by tag?
> 
> I post the patch again with 'Signed-off-by tag'.
> 
Applied, thanks.

Re: pdflush can't exit

[Hideo S. <sa...@de...>]

On Mon, 26 Feb 2007 18:38:46 +0900, Paul Mundt wrote:
> Yes, I think this is the right fix. Can you please provide a
> Signed-off-by tag?

I post the patch again with 'Signed-off-by tag'.

---
This patch secures a room for the pt_regs structure on the kernel thread's stack top.

Signed-off-by: Hideo Saito <saito@de...>

diff -pU8 -Naur arch/sh/kernel/process.c.org arch/sh/kernel/process.c
--- arch/sh/kernel/process.c.org	Mon Feb  5 03:44:54 2007
+++ arch/sh/kernel/process.c	Tue Feb 27 10:37:37 2007
@@ -258,20 +258,20 @@ int copy_thread(int nr, unsigned long cl
 
 	childregs = task_pt_regs(p);
 	*childregs = *regs;
 
 	if (user_mode(regs)) {
 		childregs->regs[15] = usp;
 		ti->addr_limit = USER_DS;
 	} else {
-		childregs->regs[15] = (unsigned long)task_stack_page(p) + THREAD_SIZE;
+		childregs->regs[15] = (unsigned long) childregs;
 		ti->addr_limit = KERNEL_DS;
 	}
-        if (clone_flags & CLONE_SETTLS) {
+	if (clone_flags & CLONE_SETTLS) {
 		childregs->gbr = childregs->regs[0];
 	}
 	childregs->regs[0] = 0; /* Set return value for child */
 
 	p->thread.sp = (unsigned long) childregs;
 	p->thread.pc = (unsigned long) ret_from_fork;
 
 	p->thread.ubc_pc = 0;

Re: PCMCIA broken on 2.6.21-rc

[Paul M. <le...@li...>]

On Mon, Feb 26, 2007 at 10:39:11AM +0100, Manuel Lauss wrote:
> On Mon, Feb 26, 2007 at 12:24:15PM +0900, Paul Mundt wrote:
> > On Wed, Feb 21, 2007 at 01:38:56PM +0100, Manuel Lauss wrote:
> > > Reverting "lazy dcache writeback optimizations" fixes both issues on
> > > my SH7760.
> > > 
> > Can you try to run Saito-san's test case and see if it ends up breaking
> > for you as well? And if so, whether it works for you after reverting the
> > changes?
> 
> With and without the patch reverted, the program does not report any
> failures (I let it run 15 minutes both times).
> 
Figures. I suppose this too will need to be reverted then until it's had
more testing on the older parts.. I'll queue up a reversion patch and
push it out with some other updates, so at least -rc2 should work for you.

Re: pdflush can't exit

[Paul M. <le...@li...>]

On Mon, Feb 26, 2007 at 05:43:05PM +0900, Hideo Saito wrote:
> When I run a preemptive kernel-2.6.20 for SH7780, a created
> kthread(pdflush) can not exit by do_exit() in kernel_thread_helper. I
> think that the created kthread should have a room for 'struct pt_regs'
> space on the stack top, because __switch_to() will refer to the space
> as follows using 'regs = task_pt_regs(prev)' and next condition may be
> true.
> 
Yes, I think this is the right fix. Can you please provide a
Signed-off-by tag?

Re: PCMCIA broken on 2.6.21-rc

[Manuel L. <ma...@ro...>]

Hi Paul,

On Mon, Feb 26, 2007 at 12:24:15PM +0900, Paul Mundt wrote:
> On Wed, Feb 21, 2007 at 01:38:56PM +0100, Manuel Lauss wrote:
> > Reverting "lazy dcache writeback optimizations" fixes both issues on
> > my SH7760.
> > 
> Can you try to run Saito-san's test case and see if it ends up breaking
> for you as well? And if so, whether it works for you after reverting the
> changes?

With and without the patch reverted, the program does not report any
failures (I let it run 15 minutes both times).

Thanks!

-- 
 ml.

pdflush can't exit

[Hideo S. <sa...@de...>]

Hi Paul,

When I run a preemptive kernel-2.6.20 for SH7780, a created kthread(pdflush) can not exit by do_exit() in kernel_thread_helper. I think that the created kthread should have a room for 'struct pt_regs' space on the stack top, because __switch_to() will refer to the space as follows using 'regs = task_pt_regs(prev)' and next condition may be true.

struct task_struct *__switch_to(struct task_struct *prev, struct task_struct *next)
{
#if defined(CONFIG_SH_FPU)
	unlazy_fpu(prev, task_pt_regs(prev));
#endif

#ifdef CONFIG_PREEMPT
	{
		unsigned long flags;
		struct pt_regs *regs;

		local_irq_save(flags);
		regs = task_pt_regs(prev);
		if (user_mode(regs) && regs->regs[15] >= 0xc0000000) {
			int offset = (int)regs->regs[15];

			/* Reset stack pointer: clear critical region mark */
			regs->regs[15] = regs->regs[1];
			if (regs->pc < regs->regs[0])
				/* Go to rewind point */
				regs->pc = regs->regs[0] + offset;
		}

I think that the following patch can avoid this problem.

--- arch/sh/kernel/process.c2	Mon Feb  5 18:52:29 2007
+++ arch/sh/kernel/process.c	Mon Feb 26 12:35:17 2007
@@ -270,20 +270,20 @@ int copy_thread(int nr, unsigned long cl
 
 	childregs = task_pt_regs(p);
 	*childregs = *regs;
 
 	if (user_mode(regs)) {
 		childregs->regs[15] = usp;
 		ti->addr_limit = USER_DS;
 	} else {
-		childregs->regs[15] = (unsigned long)task_stack_page(p) + THREAD_SIZE;
+		childregs->regs[15] = (unsigned long)childregs;
 		ti->addr_limit = KERNEL_DS;
 	}
-        if (clone_flags & CLONE_SETTLS) {
+	if (clone_flags & CLONE_SETTLS) {
 		childregs->gbr = childregs->regs[0];
 	}
 	childregs->regs[0] = 0; /* Set return value for child */
 
 	p->thread.sp = (unsigned long) childregs;
 	p->thread.pc = (unsigned long) ret_from_fork;
 
 	p->thread.ubc_pc = 0;

Re: PCMCIA broken on 2.6.21-rc

[Paul M. <le...@li...>]

Hi Manuel,

On Wed, Feb 21, 2007 at 01:38:56PM +0100, Manuel Lauss wrote:
> Reverting "lazy dcache writeback optimizations" fixes both issues on
> my SH7760.
> 
Can you try to run Saito-san's test case and see if it ends up breaking
for you as well? And if so, whether it works for you after reverting the
changes?

Re: Why flush cache in insw/outsw ?

[Nobuhiro I. <he...@t-...>]

Hi , Paul .

On Tue, 20 Feb 2007 09:50:58 +0900
Paul Mundt <le...@li...> wrote:

> 
> Iwamatsu-san, are you able to add shmin in to your build scripts? It
> would be good to get at least one SH-3 (and an SH-2) in to the nightly
> builds.

OK , I will add shmin(SH3-machine) to  nightly build .
I think that after the matter of SH2 prepares the compiler of SH2.

regards,
 Nobuhiro 
-- 
Nobuhiro Iwamatsu
	E-Mail : he...@t-...
	GPG ID : 3170EBE9 

Re: Patches for HP6xx platform

[Kristoffer E. <kri...@ho...>]

My boot problems still eludes me, my knowledge of how the offchip IRQ's are 
handled are not good enough currently. It doesnt seem to detect the hd64461 
CF card and thus is unable to boot from it.

Concerning keyboard:
I would be very much interested in that, so I can get a good idea what is 
needed.

Best wishes
Kristoffer

>From: Paul Mundt <le...@li...>
>To: Kristoffer Ericson <kri...@ho...>
>CC: lin...@li...
>Subject: Re: Patches for HP6xx platform
>Date: Fri, 23 Feb 2007 18:45:38 +0900
>
>On Wed, Feb 21, 2007 at 05:25:45PM +0000, Kristoffer Ericson wrote:
> > Btw, usually I use git diff to create patches (this time I used diff). 
>Im
> > guessing that git diff is preferred when sending you patches then?
> >
>Yes, git diff will do fine if you've got a working tree. If you've
>already checked things in locally then you can simply use something like
>'git format-patch --pretty=email <hash>'.
>
>On Wed, Feb 21, 2007 at 11:18:22PM +0000, Kristoffer Ericson wrote:
> > Paul, any idea when the keyboard for hp6xx will be pushed into git?
> >
>There are some outstanding issues with it, I prepared a patch for Dmitry
>but it still requires some attention. Someone with a vested interest in
>hp6xx will have to take it the rest of the way, I can provide the patch
>and comment mail off-list to whoever is interested.

_________________________________________________________________
Don't just search. Find. Check out the new MSN Search! 
http://search.msn.click-url.com/go/onm00200636ave/direct/01/

Re: Patches for HP6xx platform

[Paul M. <le...@li...>]

On Wed, Feb 21, 2007 at 05:25:45PM +0000, Kristoffer Ericson wrote:
> Btw, usually I use git diff to create patches (this time I used diff). Im 
> guessing that git diff is preferred when sending you patches then?
> 
Yes, git diff will do fine if you've got a working tree. If you've
already checked things in locally then you can simply use something like 
'git format-patch --pretty=email <hash>'.

On Wed, Feb 21, 2007 at 11:18:22PM +0000, Kristoffer Ericson wrote:
> Paul, any idea when the keyboard for hp6xx will be pushed into git?
> 
There are some outstanding issues with it, I prepared a patch for Dmitry
but it still requires some attention. Someone with a vested interest in
hp6xx will have to take it the rest of the way, I can provide the patch
and comment mail off-list to whoever is interested.

HP6xx bootup problems

[Kristoffer E. <kri...@ho...>]

Greetings,

Gotten a few steps closer today, but although it boots it doesnt detect the 
CF card properly and hence panics on unable to locate root. I've added some 
lines to the patch, could someone give it a quick glance and see if the 
values are correct.

Best wishes
Kristoffer
-----------------------------------------------------------------------

+#define IPRx(A,N) .addr=A, .shift=0*N*-1
+#define IPRA(N)    IPRx(0xfffffee2UL,N)
+#define IPRB(N)    IPRx(0xfffffee4UL,N)
+#define IPRC(N)    IPRx(0xa4000016UL,N)
+#define IPRD(N)    IPRx(0xa4000018UL,N)
+#define IPRE(N)    IPRx(0xa400001aUL,N)
+
+static struct ipr_data sh7709_ipr_map[] = {
+             [16]      = { IPRA(15-12),2 },    /* TMU TUNI0 */
+	     [17]      = { IPRA(11-8), 4 },    /* TMU TUNI1 */
+	     [22]      = { IPRA(3-0),  2 },    /* RTC CUI   */
+	     [23 ... 26] = { IPRB(7-4), 3 },   /* SCI */
+	     [27]        = { IPRB(15-12), 2 }, /* WDT ITI */
+             [32]        = { IPRC(0),1     },  /* IRQ 0 - xxxxxxx */
+	     [33]        = { IPRC(1),1     },  /* IRQ 1 - xxxxxxx */
+	     [34]        = { IPRC(2),1     },  /* IRQ 2 - HP6xx APM */
+             [35]        = { IPRC(3),1     },  /* IRQ 3 - HP6xx TouchScreen 
*/
+	     [36]        = { IPRD(0),1     },  /* IRQ 4 - HD64461 */
+             [37]        = { IPRD(1),1     },  /* IRQ 5 - xxxxxx */
+	     [48 ... 51] = { IPRE(15-12), 7},  /* DMA */
+	     [52 ... 55] = { IPRE(11-8), 3 },  /* IRDA */
+	     [56 ... 59] = { IPRE(7-4), 3 },   /* SCIF */
+};
+
+
+void __init init_IRQ_ipr()
+{
+  make_ipr_irq(sh7709_ipr_map, ARRAY_SIZE(sh7709_ipr_map));
+}
+

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

Implementation of HP6xx keyboard driver?

[Kristoffer E. <kri...@ho...>]

Greetings,

Paul, any idea when the keyboard for hp6xx will be pushed into git?

Best wishes
Kristoffer Ericson

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

HP6xx - vanilla 2.6.20 compiles "nicely"

[Kristoffer E. <kri...@ho...>]

Greetings,

Applied this patch 
(http://thread.gmane.org/gmane.linux.ports.sh.devel/1961/focus=1962) which 
was alot more elegant solution than mine. Kernel compiles now.

Best wishes
Kristoffer Ericson

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar - get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/

HP6xx compilation issues 2.6.20 vanilla

[Kristoffer E. <kri...@ho...>]

Greetings,

Made use of our new jlime bugtracker 
(http://jlime.com/bugtracker/main_page.php) to try and sort out the current 
issues with the 2.6.20 kernel. Just for future reference, here is a almost 
complete list of what needs doing.

Im still having issues with the ": undefined reference to `make_ipr_irq'" 
that occurs just before finishing the compile. Paul, do I need to create a 
proper ipr_map? And I will send in a minor patch to fix the get_clk changes.

Best wishes
Kristoffer Ericson
www.jlime.com

--- linux-2.6.20-vanilla/arch/sh/kernel/cpu/irq/pint.c	2007-02-04 
18:44:54.000000000 +0000
+++ linux-2.6.20/arch/sh/kernel/cpu/irq/pint.c	2007-02-21 21:12:38.000000000 
+0000
@@ -18,6 +18,38 @@
#include <asm/io.h>
#include <asm/machvec.h>

+#define INTC_IRR0   0xa4000004UL
+#define INTC_IRR1   0xa4000006UL
+#define INTC_IRR2   0xa4000008UL
+#define INTC_IPRA   0xfffffee2UL
+#define INTC_IPRB   0xfffffee4UL
+
+#define INTC_ICR0   0xfffffee0UL
+#define INTC_ICR1   0xa4000010UL
+#define INTC_ICR2   0xa4000012UL
+
+#define INTC_IPRD   0xa4000018UL
+#define INTC_INTER  0xa4000014UL
+
+#define PORT_PACR   0xa4000100UL
+#define PORT_PBCR   0xa4000102UL
+#define PORT_PCCR   0xa4000104UL
+#define PORT_PFCR   0xa400010aUL
+#define PORT_PADR   0xa4000120UL
+#define PORT_PBDR   0xa4000122UL
+#define PORT_PCDR   0xa4000124UL
+#define PORT_PFDR   0xa400012aUL
+
+#define PINT0_IRQ   40
+#define PINT8_IRQ   41
+#define PINT0_IPR_ADDR INTC_IPRD
+#define PINT8_IPR_ADDR INTC_IPRD
+
+#define PINT0_IPR_POS 3
+#define PINT8_IPR_POS 2
+#define PINT0_PRIORITY 2
+#define PINT8_PRIORITY 2
+
static unsigned char pint_map[256];
static unsigned long portcr_mask;

--- linux-2.6.20-vanilla/arch/sh/boards/hp6xx/hp6xx_apm.c	2007-02-04 
18:44:54.000000000 +0000
+++ linux-2.6.20/arch/sh/boards/hp6xx/hp6xx_apm.c	2007-02-21 
21:22:57.000000000 +0000
@@ -25,6 +25,8 @@
#define HP680_BATTERY_MIN		600
#define HP680_BATTERY_AC_ON		900

+#define IRQ0_IRQ  32
+
#define MODNAME "hp6x0_apm"

static int hp6x0_apm_get_info(char *buf, char **start, off_t fpos, int 
length)

--- linux-2.6.20-vanilla/arch/sh/cchips/hd6446x/hd64461/setup.c	2007-02-04 
18:44:54.000000000 +0000
+++ linux-2.6.20/arch/sh/cchips/hd6446x/hd64461/setup.c	2007-02-21 
21:29:16.000000000 +0000
@@ -15,6 +15,8 @@
#include <asm/irq.h>
#include <asm/hd64461.h>

+#define INTC_ICR1 0xa4000010UL
+
static void disable_hd64461_irq(unsigned int irq)
{
	unsigned short nimr;

--- 
linux-2.6.20-vanilla/drivers/input/touchscreen/hp680_ts_input.c	2007-02-04 
18:44:54.000000000 +0000
+++ linux-2.6.20/drivers/input/touchscreen/hp680_ts_input.c	2007-02-21 
21:38:03.000000000 +0000
@@ -17,11 +17,12 @@

#define	PHDR	0xa400012e
#define SCPDR	0xa4000136
+#define IRQ3_IRQ 35

static void do_softint(void *data);

static struct input_dev *hp680_ts_dev;
-static DECLARE_WORK(work, do_softint, 0);
+static DECLARE_WORK(work, do_softint);

static void do_softint(void *data)
{

--- linux-2.6.20-vanilla/sound/oss/sh_dac_audio.c	2007-02-04 
18:44:54.000000000 +0000
+++ linux-2.6.20/sound/oss/sh_dac_audio.c	2007-02-21 21:49:37.000000000 
+0000
@@ -31,6 +31,8 @@

#define MODNAME "sh_dac_audio"

+#define TIMER1_IRQ 16
+
#define TMU_TOCR_INIT	0x00

#define TMU1_TCR_INIT	0x0020	/* Clock/4, rising edge; interrupt on */
@@ -104,7 +106,7 @@
	unsigned long interval;
  	struct clk *clk;

- 	clk = clk_get("module_clk");
+ 	clk = clk_get(NULL, "module_clk");
  	interval = (clk_get_rate(clk) / 4) / rate;
  	clk_put(clk);
	ctrl_outl(interval, TMU1_TCOR);

--- linux-2.6.20-vanilla/arch/sh/drivers/dma/dma-sh.c	2007-02-04 
18:44:54.000000000 +0000
+++ linux-2.6.20/arch/sh/drivers/dma/dma-sh.c	2007-02-21 21:56:32.000000000 
+0000
@@ -19,6 +19,15 @@
#include <asm/io.h>
#include "dma-sh.h"

+#define INTC_IPRE  0xa400001aUL
+#define DMTE0_IRQ  48
+#define DMTE1_IRQ  49
+#define DMTE2_IRQ  50
+#define DMTE3_IRQ  51
+#define DMTE4_IRQ  76
+#define DMA_IPR_ADDR INTC_IPRE
+#define DMA_IPR_POS  3
+#define DMA_PRIORITY 7


#ifdef CONFIG_CPU_SH4

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

PCMCIA broken on 2.6.21-rc

[Manuel L. <ma...@ro...>]

Hi Paul,

PCMCIA is (again) broken, this time on 2.6.21-rc1.
After doing a p3_ioremap() of AREA5/6, accesses to it
will hang (a simple CIS dump for example simply hangs
on the first readb()). Kernel is not dead (sysrq-t from usb kbd works).

X breaks too, btw. It just sits there and consumes 80% CPU.

Reverting "lazy dcache writeback optimizations" fixes both issues on
my SH7760.

Thanks,

-- 
 ml.

Re: a synonym problem

[Hideo S. <sa...@de...>]

On Tue, 20 Feb 2007 10:48:12 +0900, Paul Mundt <le...@li...> wrote:
> Please also provide the n_aliases printk() from the boot log.

Feb 17 08:31:05 shlinux kernel: [42949372.970000] I-cache : n_ways=4 n_sets=256 way_incr=8192
Feb 17 08:31:05 shlinux kernel: [42949372.970000] I-cache : entry_mask=0x00001fe0 alias_mask=0x00001000 n_aliases=2
Feb 17 08:31:05 shlinux kernel: [42949372.970000] D-cache : n_ways=4 n_sets=256 way_incr=8192
Feb 17 08:31:06 shlinux kernel: [42949372.970000] D-cache : entry_mask=0x00001fe0 alias_mask=0x00001000 n_aliases=2
Feb 17 08:31:06 shlinux kernel: [42949373.190000] Mount-cache hash table entries: 512
Feb 17 08:31:06 shlinux kernel: [42949373.190000] CPU: SH7780

> Can you move your __flush_wback_region() in to the else path in
> flush_dcache_page() in arch/sh/mm/cache-sh4.c and see if the problem
> persists?

I backed out the change for install_arg_page() at fs/exec.c and applied following patch and tested it using appended test program. I can not see this problem by my test.

--- arch/sh/mm/cache-sh4.c.org	Mon Feb  5 03:44:54 2007
+++ arch/sh/mm/cache-sh4.c	Tue Feb 20 12:59:46 2007
@@ -243,16 +243,21 @@ void flush_dcache_page(struct page *page
 		unsigned long phys = PHYSADDR(page_address(page));
 		unsigned long addr = CACHE_OC_ADDRESS_ARRAY;
 		int i, n;
 
 		/* Loop all the D-cache */
 		n = cpu_data->dcache.n_aliases;
 		for (i = 0; i < n; i++, addr += 4096)
 			flush_cache_4096(addr, phys);
+	} else {
+		unsigned long phys = PHYSADDR(page_address(page));
+		int size = cpu_data->dcache.n_aliases << PAGE_SHIFT;
+
+		__flush_wback_region((void *)P1SEGADDR(phys), size);
 	}
 
 	wmb();
 }
 
 /* TODO: Selective icache invalidation through IC address array.. */
 static inline void flush_icache_all(void)
 {

-error log if original code---------------------------
% ./vfork_test 
******29 != 30
*stat = 0x100

-test programs----------------------------------------
% gcc -o args args.c
% gcc -o vfork_test vfork_test.c
-args.c-----------------------------------------------
#include <stdio.h>
#include <stdlib.h>

int main(int ac, char **av)
{
	int n = 1;
	int nn = 0;
	int err = 0;

	ac--, av++;
	while (ac > 0) {
		nn = atoi(*av);
		if (nn != n) {
			fprintf(stderr, "%d != %d\n", nn , n);
			err++;
			break;
		}
		n++;
		ac--, av++;
	}
	fprintf(stderr, "*");
	exit(err);
}

-vfork_test.c-------------------------------------------------
#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <sys/types.h>
#include <sys/wait.h>

#define ARGS	1000

char* argv[ARGS];

exec_test(int ac, char** av)
{
	int id, stat;

	id = vfork();
	if (id == 0) {
		execve("./args", argv, NULL);
		fprintf(stderr, "can not exec args command\n");
		exit(0);
	}
	wait(&stat);
	return stat;
}

int main(int ac, char** av)
{
	int i, stat;
	char* str;

	str = calloc(ARGS, 12);
	for (i = 0; i < ARGS; i++) {
		sprintf(str, "%010d", i);
		argv[i] = str;
		str += strlen(str) + 1;
	}
	for (;;) {
		stat = exec_test(ARGS, argv);
		if (WEXITSTATUS(stat) != 0) {
			printf("stat = 0x%x\n", stat);
			exit(1);
		}
	}
}

Re: Patches for HP6xx platform

[Paul M. <le...@li...>]

On Tue, Feb 20, 2007 at 01:45:57AM +0000, Kristoffer Ericson wrote:
> Finally got around to do some test builds. Didnt get very far. I've been 
> tracking down the changes and got a first set of patches.
> 
A minor nit. When submitting patches to the list, can you please use
text/plain instead of text/x-patch or something equally fancy? Some
mailers are rather fond of the latter, but replying to the patches inline
requires manual mangling of the mime type, and while I do have a macro
for this, I rather wish I didn't.

> Touchscreen patch fixes the change DECLARE_WORK(1,2,3) ->
> DECLARE_WORK(1.2) that has taken place.
> 
This looks fine. Please send a Signed-off-by tag for these.

> Pint.c requires some variables to work properly. I've taken those from the 
> old irq.h file. Im not sure if those are bound to the hp6xx platform and if 
> so where they should go.
> 
These should probably move in to the hp6xx-specific parts and be handed
off, but I'm currently overhauling some of how we do that, so you may
want to hold off on some of that and just use this patch locally until
then.

> +#define INTC_IPRD 0xffd0010UL
> +#define INTC_ICR2 0xa4000012UL
> +#define PORT_PCCR 0xA4050104UL
> +#define PORT_PCDR 0xA4050106UL
> +#define PORT_PFDR 0xa400012aUL
> +#define PORT_PFCR 0xa405010AUL
> +

IPRD is completely bogus. Copy and paste error, or was this always wrong?

Re: a synonym problem

[Paul M. <le...@li...>]

Saito-san,

On Tue, Feb 20, 2007 at 10:03:15AM +0900, Hideo Saito wrote:
> This problem is caused when a process is executed after vfork. In the
> result, the executed process can not read the arguments correctly.
> 
> I think that the copied arguments from parent process should be copied
> back to the physical memory as following patch, in other words, a
> mapped page to process should be copied back to the physical memory
> unconditionally.
> 
This is quite interesting. These are anonymous pages, so page->mapping
will be NULL unless you've got a swap cache page. The writeback should
not be deferred in this case.

Can you move your __flush_wback_region() in to the else path in
flush_dcache_page() in arch/sh/mm/cache-sh4.c and see if the problem
persists? Please also provide the n_aliases printk() from the boot log.

Patches for HP6xx platform

[Kristoffer E. <kri...@ho...>]

Greetings,

Finally got around to do some test builds. Didnt get very far. I've been 
tracking down the changes and got a first set of patches.

Paul, please give me some input on where to put the vars (if my location is 
correct).

Touchscreen patch fixes the change DECLARE_WORK(1,2,3) -> DECLARE_WORK(1.2) 
that
has taken place.

Pint.c requires some variables to work properly. I've taken those from the 
old irq.h file. Im not sure if those are bound to the hp6xx platform and if 
so where they should go.

All patches are against latest git

Best wishes
Kristoffer Ericson

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/