Something about elp I need help with.

Something about elp I need help with.

[Alan Mackenzie]

Hi, Emacs.

Whilst trying to track down what is dragging in a fontification event, I
get the following output from elp-results (blank lines added for
clarity):

Function Name                         Call Count Elapsed Time Average Time
c-beginning-of-decl-1                  90          8.767888      0.0974209777
c-beginning-of-statement-1             103         8.7664949999  0.0851116019
c-crosses-statement-barrier-p          14519       7.6718619999  0.0005284015
c-at-macro-vsemi-p                     20230       4.3577799999  0.0002154117
c-in-literal                           20230       3.8880240000  0.0001921910

c-literal-limits                       22322       2.2931949999  0.0001027325
c-backward-sws                         72797       0.9805889999  1.347...e-05
c-state-safe-place                     31416       0.3780669999  1.203...e-05
c-beginning-of-macro                   44678       0.3394430000  7.597...e-06
c-font-lock-declarations               8           0.2129779999  0.0266222499  <========

c-find-decl-spots                      8           0.2129130000  0.0266141250
c-parse-state                          177         0.1174700000  0.0006636723

All these functions down to c-font-lock-declarations are low level
functions, called only from other CC Mode ones.  Yet there don't seem to
be any HL functions calling the LL ones at the top, such as
c-beginning-of-decl-1.

How can this be?  What have I misunderstood about elp?

Thanks for the help!

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: Something about elp I need help with.

[Alan Mackenzie]

On Fri, Jan 13, 2012 at 04:03:09PM +0000, Alan Mackenzie wrote:
> Hi, Emacs.

> Whilst trying to track down what is dragging in a fontification event, I
> get the following output from elp-results (blank lines added for
> clarity):

> Function Name                         Call Count Elapsed Time Average Time
> c-beginning-of-decl-1                  90          8.767888      0.0974209777
> c-beginning-of-statement-1             103         8.7664949999  0.0851116019
> c-crosses-statement-barrier-p          14519       7.6718619999  0.0005284015
> c-at-macro-vsemi-p                     20230       4.3577799999  0.0002154117
> c-in-literal                           20230       3.8880240000  0.0001921910

> c-literal-limits                       22322       2.2931949999  0.0001027325
> c-backward-sws                         72797       0.9805889999  1.347...e-05
> c-state-safe-place                     31416       0.3780669999  1.203...e-05
> c-beginning-of-macro                   44678       0.3394430000  7.597...e-06
> c-font-lock-declarations               8           0.2129779999  0.0266222499  <========

> c-find-decl-spots                      8           0.2129130000  0.0266141250
> c-parse-state                          177         0.1174700000  0.0006636723

> All these functions down to c-font-lock-declarations are low level
> functions, called only from other CC Mode ones.  Yet there don't seem to
> be any HL functions calling the LL ones at the top, such as
> c-beginning-of-decl-1.

> How can this be?  What have I misunderstood about elp?

I've worked it out, now.  There was a macro generated function called
directly from font-lock.  It was the "missing" high level function.

> Thanks for the help!

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: Something about elp I need help with.

[Stefan Monnier]

>> How can this be?  What have I misunderstood about elp?
> I've worked it out, now.

I'm glad we could help,


        Stefan ;-)