On 29/09/14 18:21, ph10@??? wrote:
> In the new API, there is no longer an explicit "study" function - as it
> turned out to be cheap, it now always done. For JIT there will be a new
> function called pcre2_jit_compile(), to be called after pcre2_compile()
> has been successful. Freeing of the JIT memory will happen automatically
> when pcre2_code_free() is called.
sounds good.
> It would be relatively simple to add support for (*JIT) to PCRE2
Sounds good, too.
> However, I am slightly wondering why you need to offer the option to
> varnish users rather than just enabling JIT always, in order to get the
> fastest performance. What have I missed?
There had been issues which I hadn't been involed with personally at the time -
see
https://www.varnish-cache.org/trac/ticket/1191
Enabling JIT has been possible since then through a global compile time
(configure) option. I do not have information about how wide spread use of this
option is at the moment.
So far I have not been involved deeply enough with the pcre project to identify
potential known issues which could have root caused this and neither am I aware
of anyone else from the varnish dev community who is, so pointers so relevant
known issues are welcome.
Would you (safely) advise to turn on JIT by default for some minimum pcre version?
Nils
