Hi Tom,

Unfortunately, it's not possible to just revert the removal of no-VFP. We
now require a subset of the Crankshaft compiler for all platforms that we
support, and supporting no-VFP Crankshaft code was simply too much work for
the value added (Chrome for Android is not support on any of those devices)
and maintaining the no-VFP code slowed our progress on the core supported
platforms.

I don't think that we even would want to bring no-VFP back to trunk V8,
even if you were to magically provide full no-VFP support including
Crankshaft based on bleeding edge in a patch for us. The complexity the
port adds still outweighs the benefit going forward. If your crash is
indeed due to missing VFP support, you will have to use an older version of
V8 for those devices.

Regards,
Danno


On Wed, Oct 30, 2013 at 12:58 PM, Rodolph Perfetta <
[email protected]> wrote:

> Hi Tom,
>
> Can you confirm this is a VFP issue:
>  * adb shell cat /proc/cpuinfo on the device will tell you if you have VFP
>  * what instruction is generating the fault?
>
> On a side note reverting the commit on a post 3.18 V8 will not work.
>
> Cheers,
> Rodolph.
>
>
>
> On 30 October 2013 11:35, Tom Jacobs <[email protected]> wrote:
>
>> Hi Danno,
>>
>> We're running V8 on Android and we're finding that it's crashing on some
>> older Android devices (a few years old), we suspect those that don't have
>> VFPv2.
>> I see support for no-VFPv2 was removed in 3.18 in this commit:
>> https://code.google.com/p/v8/source/detail?r=14159
>>
>> Is there any way of bringing no-VFPv2 support back to V8? We're running a
>> fork of it from about 6 months ago (https://github.com/Expensify/jv8),
>> and so might it be as simple as just reverting that commit?
>>
>> Or is there any other way of working around and having it run fine on
>> those chips?
>>
>> Our crash happens here:
>> Trying to load lib /data/data/org.me.**mobiexpensifyg/lib/libjv8.so
>> 0x46030c28
>> ... crash ...
>> #00  pc *0023c978*  /data/data/org.me.**mobiexpensifyg/lib/libjv8.so
>> By analyzing which line of code the crash comes from, this is what I get:
>> $ addr2line -C -f -e obj/local/armeabi/libjv8.so *0023c978*
>> v8::internal::Isolate::**EnsureDefaultIsolate()
>>
>> Thanks,
>> Tom.
>>
>>
>> On Thursday, April 4, 2013 10:52:22 PM UTC+11, danno wrote:
>>>
>>> Due the increasing cost of the keeping the "no-VFPv2" port of V8 working
>>> on ARM, we are planning on making 3.17 the last V8 release that that
>>> supports ARM chips without VFPv2. Starting with the 3.18 release, the
>>> minimal V8 requirements will increase to ARMv6 + VFPv2. In order to
>>> simplify maintenance, we will also remove the "pre-VFP2" ARM code from the
>>> V8 code base.
>>>
>>> Danno
>>>
>>  --
>> --
>> v8-users mailing list
>> [email protected]
>> http://groups.google.com/group/v8-users
>> ---
>> You received this message because you are subscribed to the Google Groups
>> "v8-users" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected].
>> For more options, visit https://groups.google.com/groups/opt_out.
>>
>
>  --
> --
> v8-users mailing list
> [email protected]
> http://groups.google.com/group/v8-users
> ---
> You received this message because you are subscribed to the Google Groups
> "v8-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/groups/opt_out.
>

-- 
-- 
v8-users mailing list
[email protected]
http://groups.google.com/group/v8-users
--- 
You received this message because you are subscribed to the Google Groups 
"v8-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to