Hi, 
thanks for your help but I don't think to be solution.
I had apply this patch :
http://lists.infradead.org/pipermail/linux-arm-kernel/2010-December/036055.html
You can read the first problems here:
http://lists.linaro.org/pipermail/linaro-kernel/2014-March/012635.html

I had found a solution to this problem, and I write to you into my first e-mail.
Now I have this problems:
When resume  and run this function
+ENTRY(swsusp_arch_resume)
+	/*
+	 * Restore_pblist is the starting point for loaded pages
+	 */
+	ldr     r0, .Lrestore_pblist
+	ldr     r6, [r0]
+
+.Lcopy_loop:
+	ldr     r4, [r6]     /* src IOW present address */
+	ldr     r5, [r6, #4] /* dst IOW original address*/
+	mov     r9, #1024    /* No. of entries in one page, where each entry is 4 bytes */
+
+.Lcopy_one_page:
+	/*
+	 * This loop could be optimized by using stm and ldm.
+	 */
+	ldr     r8, [r4], #4
+	str     r8, [r5], #4
+	subs    r9, r9, #1
+	bne     .Lcopy_one_page
+
+	/* The last field of struct pbe is a pointer to the next pbe structure */
+	ldr     r6, [r6, #8]
+	cmp     r6, #0
+	bne     .Lcopy_loop
+
+	/*
+	 * Restore SVC context
+	 */
+	ldr     r3, .Lsaved_context_r13_svc
+	ldmia   r3, {r13-r14}
+	ldr     r3, .Lsaved_spsr_svc
+	ldr     r1, [r3]
+	msr     spsr_cxsf, r1
+
+	mrs     r0, cpsr	/* Save current mode into r0 */
+
+	/*
+	 * Change to system(user) mode
+	 */
+	mov     r1, r0
+	orr     r1, r1, #0x1f
+	msr     cpsr_c, r1
+
+	/*
+	 * Restore User context
+	 */
+	ldr     r3, .Lsaved_context_r0
+	ldmia   r3, {r0-r14}
+	ldr     r3, .Lsaved_cpsr
+	ldr     r1, [r3]
+	msr     cpsr_cxsf, r1
+
+	msr     cpsr_c, r0	/* Restore original mode from r0 */
+
+	/*
+	 * Flush TLB (Invalidate unified TLB unlocked entries)
+	 */
+	mov     r1, #0
+	mcr     p15, 0, r1, c8, c7, 0
+
+	/* Set the return value */
+	mov	r0, #0
+
+	/* Restore return address */
+	ldr     r3, .Lsaved_context_r14_svc
+	ldr     lr, [r3]
+	mov     pc, lr
+ENDPROC(swsusp_arch_resume)

After about 700/800 loop the function stop to work in BOLD istruction.

Have you idea for solution?

Thanks


Fioretti Luigi
mailto:luigi.fioretti@dinema.it
Skype: Dinema.Fioretti.Luigi
Tel : 00390302300492
Dinema Spa

Via S.Polo 183
25124 Brescia


-----linaro-kernel-bounces@lists.linaro.org ha scritto: -----
Per: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Da: Viresh Kumar
Inviato da: linaro-kernel-bounces@lists.linaro.org
Data: 05/01/2015 05.05AM
Cc: Prarit Bhargava <prarit@redhat.com>, Lists linaro-kernel <linaro-kernel@lists.linaro.org>, Saravana Kannan <skannan@codeaurora.org>, "linux-pm@vger.kernel.org" <linux-pm@vger.kernel.org>
Oggetto: Re: [PATCH V2 04/14] cpufreq: stats: pass 'stat' to cpufreq_stats_update() *

On 4 January 2015 at 03:34, Rafael J. Wysocki <rjw@rjwysocki.net> wrote:
> What about:
>
> "It is better to pass a struct cpufreq_stats pointer to cpufreq_stats_update()
>  instead of a CPU number, because ..."

Sure.

>> +static int cpufreq_stats_update(struct cpufreq_stats *stat)
>>  {
>> -     struct cpufreq_stats *stat;
>> -     unsigned long long cur_time;
>> +     unsigned long long cur_time = get_jiffies_64();
>>
>> -     cur_time = get_jiffies_64();
>
> Unrelated change.

Yeah, it was too small of a change so not that significant as well.
And so folded it into this commit only :(

Will send it separately in v3..

_______________________________________________
linaro-kernel mailing list
linaro-kernel@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-kernel