On Tuesday, August 11, 2015 at 1:44:42 PM UTC+2, Yang Guo wrote:
>
> Hi,
>
> which V8 version are you using? I don't think you can hit that assertion 
> with the code serializer with the most up-to-date V8. kProduceParserCache 
> doesn't produce any cached code, just some hint to speed up parsing. In 
> some cases there is no hint that the parser can take advantage of.
>

I was using version 3.28.73 which was the version released with Node at 
that point. I tried switching to 4.3 and yes the problem went away and the 
serializer now seems to work. Thanks.

 

> What you should instead use is the custom start-up snapshot. You would 
> create a snapshot of the heap after you executed the startup script, so 
> that later, starting up would already put the heap into the state you 
> snapshotted. Here's a short explaination: 
> http://www.hashseed.net/2015/03/improving-v8s-performance-using.html
>

That seems like a great idea. However, I am not sure I understand exactly 
how that would work. Can I only pass one script? For instance, if I know I 
need all the NodeJs lib/*.js files, would I be able to capture the state of 
all of those, or can I only pick one script? I read somewhere that the blob 
will not work with node.js require function as that one is implemented in 
native C++.

 

-- 
-- 
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/d/optout.

Reply via email to