[Qemu-devel] [PATCH] icount: Fix virtual clock start value on ARM

[Qemu-devel] [PATCH] icount: Fix virtual clock start value on ARM

[Sebastian Tanase]

When using the icount option on ARM, the virtual
clock starts counting at realtime clock but it
should start at 0.
This small fix addresses this issue.

Signed-off-by: Sebastian Tanase <sebastian.tanase@openwide.fr>
---
 cpus.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/cpus.c b/cpus.c
index 5e7f2cf..de18ece 100644
--- a/cpus.c
+++ b/cpus.c
@@ -104,7 +104,7 @@ static bool all_cpu_threads_idle(void)
 
 /* Compensate for varying guest execution speed.  */
 static int64_t qemu_icount_bias;
-static int64_t vm_clock_warp_start;
+static int64_t vm_clock_warp_start = -1;
 /* Conversion factor from emulated instructions to virtual clock ticks.  */
 static int icount_time_shift;
 /* Arbitrarily pick 1MIPS as the minimum allowable speed.  */
-- 
2.0.0.rc2

Re: [Qemu-devel] [PATCH] icount: Fix virtual clock start value on ARM

[Paolo Bonzini]

Il 23/07/2014 11:11, Sebastian Tanase ha scritto:
> When using the icount option on ARM, the virtual
> clock starts counting at realtime clock but it
> should start at 0.
> This small fix addresses this issue.
> 
> Signed-off-by: Sebastian Tanase <sebastian.tanase@openwide.fr>

Thanks, this is ok for 2.2.

Paolo

> ---
>  cpus.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/cpus.c b/cpus.c
> index 5e7f2cf..de18ece 100644
> --- a/cpus.c
> +++ b/cpus.c
> @@ -104,7 +104,7 @@ static bool all_cpu_threads_idle(void)
>  
>  /* Compensate for varying guest execution speed.  */
>  static int64_t qemu_icount_bias;
> -static int64_t vm_clock_warp_start;
> +static int64_t vm_clock_warp_start = -1;
>  /* Conversion factor from emulated instructions to virtual clock ticks.  */
>  static int icount_time_shift;
>  /* Arbitrarily pick 1MIPS as the minimum allowable speed.  */
> 

Re: [Qemu-devel] [PATCH] icount: Fix virtual clock start value on ARM

[Andreas Färber]

Am 23.07.2014 11:16, schrieb Paolo Bonzini:
> Il 23/07/2014 11:11, Sebastian Tanase ha scritto:
>> When using the icount option on ARM, the virtual
>> clock starts counting at realtime clock but it
>> should start at 0.
>> This small fix addresses this issue.
>>
>> Signed-off-by: Sebastian Tanase <sebastian.tanase@openwide.fr>
> 
> Thanks, this is ok for 2.2.

Could we get an explanation (in the commit message) of why this fixes
that issue? :) By my reading -1 != 0.

Thanks,
Andreas

-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg

Re: [Qemu-devel] [PATCH] icount: Fix virtual clock start value on ARM

[Peter Maydell]

On 23 July 2014 10:11, Sebastian Tanase <sebastian.tanase@openwide.fr> wrote:
> When using the icount option on ARM, the virtual
> clock starts counting at realtime clock but it
> should start at 0.
> This small fix addresses this issue.
>
> Signed-off-by: Sebastian Tanase <sebastian.tanase@openwide.fr>
> ---
>  cpus.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/cpus.c b/cpus.c
> index 5e7f2cf..de18ece 100644
> --- a/cpus.c
> +++ b/cpus.c
> @@ -104,7 +104,7 @@ static bool all_cpu_threads_idle(void)
>
>  /* Compensate for varying guest execution speed.  */
>  static int64_t qemu_icount_bias;
> -static int64_t vm_clock_warp_start;
> +static int64_t vm_clock_warp_start = -1;
>  /* Conversion factor from emulated instructions to virtual clock ticks.  */
>  static int icount_time_shift;
>  /* Arbitrarily pick 1MIPS as the minimum allowable speed.  */

Commit message says this is fixing an ARM bug but this is
a generic file. Is this actually a bug with wider scope than just
ARM?

thanks
-- PMM

Re: [Qemu-devel] [PATCH] icount: Fix virtual clock start value on ARM

[Paolo Bonzini]

Il 23/07/2014 11:41, Peter Maydell ha scritto:
> On 23 July 2014 10:11, Sebastian Tanase <sebastian.tanase@openwide.fr> wrote:
>> When using the icount option on ARM, the virtual
>> clock starts counting at realtime clock but it
>> should start at 0.
>> This small fix addresses this issue.
>>
>> Signed-off-by: Sebastian Tanase <sebastian.tanase@openwide.fr>
>> ---
>>  cpus.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/cpus.c b/cpus.c
>> index 5e7f2cf..de18ece 100644
>> --- a/cpus.c
>> +++ b/cpus.c
>> @@ -104,7 +104,7 @@ static bool all_cpu_threads_idle(void)
>>
>>  /* Compensate for varying guest execution speed.  */
>>  static int64_t qemu_icount_bias;
>> -static int64_t vm_clock_warp_start;
>> +static int64_t vm_clock_warp_start = -1;
>>  /* Conversion factor from emulated instructions to virtual clock ticks.  */
>>  static int icount_time_shift;
>>  /* Arbitrarily pick 1MIPS as the minimum allowable speed.  */
> 
> Commit message says this is fixing an ARM bug but this is
> a generic file. Is this actually a bug with wider scope than just
> ARM?

Yes, see the discussion yesterday under "Re: [RFC PATCH V4 6/6] monitor:
Add drift info to 'info jit'" and "Re: [RFC PATCH V4 0/6] icount:
Implement delay algorithm between guest and host clocks".

Paolo