Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On 24 Nov 2015, at 6:36 PM, Yann Ylavic wrote: > Sure, but my point is that the worst case is likely depend on the > application, It will depend on the application, yes. > eg: > >case 'm': >case 'M': >if (!strncmp(token, "max-age", 7) >||

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 10:18 AM, Graham Leggett wrote: > On 24 Nov 2015, at 6:15 PM, Yann Ylavic wrote: > > > Not sure: > >if (!strcmp(h, "max-age") > >|| ap_cmpcasestr(h, "max-age")) > > is likely to be a bit faster than a single

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 12:03 PM, Jim Jagielski wrote: > > > On Nov 24, 2015, at 11:18 AM, Graham Leggett wrote: > > > > On 24 Nov 2015, at 6:15 PM, Yann Ylavic wrote: > > > >> Not sure: > >> if (!strcmp(h, "max-age") > >> ||

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

> On Nov 24, 2015, at 1:11 PM, Yann Ylavic wrote: > > On Tue, Nov 24, 2015 at 7:03 PM, Jim Jagielski wrote: >> >>> On Nov 24, 2015, at 11:18 AM, Graham Leggett wrote: >>> >>> On 24 Nov 2015, at 6:15 PM, Yann Ylavic

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 7:29 PM, William A Rowe Jr wrote: > > I'm wondering the other way around. Even in Yann's latest exercise, > simply testing > >if ((*ps1 == *ps2) || (ucharmap[*ps1] != ucharmap[*ps2])) { > > (or in Yann's code, use the const int lookups,

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 5:18 PM, Graham Leggett wrote: > On 24 Nov 2015, at 6:15 PM, Yann Ylavic wrote: > >> Not sure: >>if (!strcmp(h, "max-age") >>|| ap_cmpcasestr(h, "max-age")) >> is likely to be a bit faster than a single ap_cmpcasestr()

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 5:09 PM, Graham Leggett wrote: > > A further optimisation - in many cases the fast path is really a case > sensitive string comparison, given that the vast majority of the time the > case being used is the case quoted in the spec. > > In other words,

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On 24 Nov 2015, at 6:15 PM, Yann Ylavic wrote: > Not sure: >if (!strcmp(h, "max-age") >|| ap_cmpcasestr(h, "max-age")) > is likely to be a bit faster than a single ap_cmpcasestr() when it > matches, but much slower when it does not. Yep, that’s the point. The

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

> On Nov 24, 2015, at 11:18 AM, Graham Leggett wrote: > > On 24 Nov 2015, at 6:15 PM, Yann Ylavic wrote: > >> Not sure: >> if (!strcmp(h, "max-age") >> || ap_cmpcasestr(h, "max-age")) >> is likely to be a bit faster than a single ap_cmpcasestr()

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

+1 (concept) > On Nov 24, 2015, at 12:40 PM, William A Rowe Jr wrote: > > On Tue, Nov 24, 2015 at 10:18 AM, Graham Leggett wrote: > On 24 Nov 2015, at 6:15 PM, Yann Ylavic wrote: > > > Not sure: > >if (!strcmp(h, "max-age") > >

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On 24 Nov 2015, at 5:21 PM, Jim Jagielski wrote: > For these types of paths, where we use a switch/case against > the 1st char of the token, the real reason why we used that > impl was to avoid the (expected) expensive string comparison. > > This is really no longer an issue.

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 7:03 PM, Jim Jagielski wrote: > >> On Nov 24, 2015, at 11:18 AM, Graham Leggett wrote: >> >> On 24 Nov 2015, at 6:15 PM, Yann Ylavic wrote: >> >>> Not sure: >>> if (!strcmp(h, "max-age") >>> ||

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

For these types of paths, where we use a switch/case against the 1st char of the token, the real reason why we used that impl was to avoid the (expected) expensive string comparison. This is really no longer an issue. Sure, we could still use this "fast path" short-circuit, but even that is

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 10:21 AM, Jim Jagielski wrote: > For these types of paths, where we use a switch/case against > the 1st char of the token, the real reason why we used that > impl was to avoid the (expected) expensive string comparison. Maybe naively, I thought we were

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 4:28 PM, Eric Covener wrote: > On Tue, Nov 24, 2015 at 10:21 AM, Jim Jagielski wrote: >> For these types of paths, where we use a switch/case against >> the 1st char of the token, the real reason why we used that >> impl was to avoid

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

> On Nov 24, 2015, at 10:28 AM, Eric Covener wrote: > > On Tue, Nov 24, 2015 at 10:21 AM, Jim Jagielski wrote: >> For these types of paths, where we use a switch/case against >> the 1st char of the token, the real reason why we used that >> impl was to

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 9:28 AM, Eric Covener wrote: > On Tue, Nov 24, 2015 at 10:21 AM, Jim Jagielski wrote: > > For these types of paths, where we use a switch/case against > > the 1st char of the token, the real reason why we used that > > impl was to

Re: svn commit: r1715938 - /httpd/httpd/trunk/modules/cache/cache_util.c

On Tue, Nov 24, 2015 at 4:49 PM, William A Rowe Jr wrote: > > Does it make sense to expose the ap_tolower_ascii() function? +1, possibly an inline one or a macro... > We could call it ap_tolower_plaintext() I prefer ascii().