Yes, that part is expected - the first time you run after an update, it
will rebuild the standard library and such. But then it caches the result.

If any of you see consistent slowdowns after that initial run, please
either bisect, or provide a bitcode file and steps to reproduce and I can
bisect.

- Alon


On Thu, Oct 9, 2014 at 4:56 PM, Brion Vibber <[email protected]> wrote:

> I'm seeing a more modest slowdown with 1.25.0 vs 1.22.0, with the final
> emcc invocation on my ogv.js library taking ~10s instead of ~4s. This is on
> a 2012 Retina MacBook Pro running 10.9.5.
>
> However I also notice a *much* larger slowdown the first time I run it
> after switching to each version, when it says "running sanity checks"...
> after that first run I get more consistent results (the 10s vs 4s), but it
> can take an extra 40-100 seconds to do the sanity checks.
>
> -- brion vibber (brion @ pobox.com / bvibber @ wikimedia.org)
>
>
> On Thu, Oct 9, 2014 at 4:34 PM, Olivier Chatry <[email protected]>
> wrote:
>
>> Hello,
>>
>> Yep I'm seeing the exact same behavior on three macs : one vm, one mini
>> mac, and my 15" retina with "latest" version. I'm not really sure what is
>> going on, I will try to get a master version on the retina see if it is
>> better.
>>
>> On Thursday, October 9, 2014 8:16:11 PM UTC+2, Alon Zakai wrote:
>>>
>>> I can't think of any relevant changes, so we should try to rule out
>>> other explanations. Is the only difference the emscripten version,
>>> otherwise it's the same command, same input file?
>>>
>>> Do you see the same effect on your other machine? If so, can you provide
>>> the input file so we can test?
>>>
>>> - Alon
>>>
>>> On Thu, Oct 9, 2014 at 5:43 AM, Olivier Chatry <[email protected]>
>>> wrote:
>>>
>>>> Seems like :
>>>>
>>>> DEBUG    root: emcc step "bitcodeize inputs" took 12.47 seconds
>>>>
>>>> is taking most of the time ( it's on i7 4770, on our build machine,
>>>> it's more like 1 minutes for a file :/ ).
>>>>
>>>> On Thursday, October 9, 2014 12:41:40 PM UTC+2, jj wrote:
>>>>>
>>>>> Try setting the environment variable EMCC_DEBUG=1 before building to
>>>>> investigate which part has slowed down. Perhaps that will give a clue to
>>>>> pinning down where it regressed?
>>>>>
>>>>> 2014-10-09 2:16 GMT-04:00 Olivier Chatry <[email protected]>:
>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> I updated emscripten yesterday, on my OSx computer, and since then
>>>>>> the build is very very slow :(. I have try the latest on linux ( which
>>>>>> seems to be "master" ) and everything is working fine.
>>>>>>
>>>>>> Any idea on how to make it faster ? or should I just revert to the
>>>>>> old version ?
>>>>>>
>>>>>> --
>>>>>> You received this message because you are subscribed to the Google
>>>>>> Groups "emscripten-discuss" 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/d/optout.
>>>>>>
>>>>>
>>>>>  --
>>>> You received this message because you are subscribed to the Google
>>>> Groups "emscripten-discuss" 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/d/optout.
>>>>
>>>
>>>  --
>> You received this message because you are subscribed to the Google Groups
>> "emscripten-discuss" 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/d/optout.
>>
>
>  --
> You received this message because you are subscribed to the Google Groups
> "emscripten-discuss" 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/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"emscripten-discuss" 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/d/optout.

Reply via email to