git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Ramsay Jones <ramsay@ramsayjones.plus.com>
To: Derrick Stolee <stolee@gmail.com>, Junio C Hamano <gitster@pobox.com>
Cc: GIT Mailing-list <git@vger.kernel.org>
Subject: Re: [PATCH 9/9] commit-reach.h: add missing declarations (hdr-check)
Date: Thu, 20 Sep 2018 16:35:16 +0100	[thread overview]
Message-ID: <2809251f-6eba-b0ac-68fe-b972809ccff7@ramsayjones.plus.com> (raw)
In-Reply-To: <eb77d563-8b3e-9b00-59b8-b050e6378c69@gmail.com>



On 20/09/18 00:38, Derrick Stolee wrote:
> On 9/18/2018 8:15 PM, Ramsay Jones wrote:
>> Signed-off-by: Ramsay Jones <ramsay@ramsayjones.plus.com>
>> ---
>>   commit-reach.h | 5 +++--
>>   1 file changed, 3 insertions(+), 2 deletions(-)
>>
>> diff --git a/commit-reach.h b/commit-reach.h
>> index 7d313e2975..f41d8f6ba3 100644
>> --- a/commit-reach.h
>> +++ b/commit-reach.h
>> @@ -1,12 +1,13 @@
>>   #ifndef __COMMIT_REACH_H__
>>   #define __COMMIT_REACH_H__
>>   +#include "commit.h"
>>   #include "commit-slab.h"
>>   -struct commit;
>>   struct commit_list;
>> -struct contains_cache;
> 
> Interesting that you needed all of commit.h here, and these 'struct commit' and 'struct contains_cache' were not enough. Was there a reason you needed the entire header file instead of just adding a missing struct declaration?

Actually, the original version of this patch didn't add that
include! I changed my mind just before sending this series
out, since I felt the original patch was conflating two issues.

The reason for the #include of 'commit.h' in this patch, but
not with my original patch, has to to with the inline functions
used in the commit-slab implementation. My original patch used
the 'commit-slab-{decl,impl}.h' header files to sidestep the
need for the definition of 'struct commit'.

I have included an 'RFC on-top' patch below to show you what I
had in mind. NOTE: I had not finished writing the commit message
and you could argue that the 'implement' macro should go in the
ref-filter.c file, since that is were the contains_cache is 
'defined and initialised'. I had not intended to send this to
the list ... :-D

Note that, if you compile with optimizations disabled, then
run this script:

  $ cat -n dup-static.sh
       1 #!/bin/sh
       2 
       3 nm $1 | grep ' t ' | cut -d' ' -f3 | sort | uniq -c |
       4 	sort -rn | grep -v '      1'
  $ 

  $ ./dup-static.sh git | grep contains
       24 init_contains_cache_with_stride
       24 init_contains_cache
       24 contains_cache_peek
       24 contains_cache_at_peek
       24 contains_cache_at
       24 clear_contains_cache
  $ 
  
you will find 24 copies of the commit-slab routines for the
contains_cache. Of course, when you enable optimizations again,
these duplicate static functions (mostly) disappear. (Two static
functions remain, the rest are actually inlined at -O2).

However, if you apply the patch below, you end up with all of
the functions in the contains_cache commit-slab implementation
as external functions. Some of those functions are never called,
so it's not necessarily a win ... ;-)

ATB,
Ramsay Jones

-- >8 --
Subject: [PATCH] commit-reach: use a shared commit-slab for the contains_cache

Commit a9f1f1f9f8 ("commit-slab.h: code split", 2018-05-19) separated
the commit-slab interface from its implementation, to allow for the
definition of a public commit-slab data structure. This enabled us to
avoid including the commit-slab implementation in a header file, which
could result in the replication of the commit-slab functions in each
compilation unit in which it was included.

Signed-off-by: Ramsay Jones <ramsay@ramsayjones.plus.com>
---
 commit-reach.c | 3 +++
 commit-reach.h | 6 +++---
 2 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/commit-reach.c b/commit-reach.c
index 622eeb313d..7223cacdd1 100644
--- a/commit-reach.c
+++ b/commit-reach.c
@@ -1,5 +1,6 @@
 #include "cache.h"
 #include "commit.h"
+#include "commit-slab-impl.h"
 #include "commit-graph.h"
 #include "decorate.h"
 #include "prio-queue.h"
@@ -18,6 +19,8 @@
 
 static const unsigned all_flags = (PARENT1 | PARENT2 | STALE | RESULT);
 
+implement_shared_commit_slab(contains_cache, enum contains_result);
+
 static int queue_has_nonstale(struct prio_queue *queue)
 {
 	int i;
diff --git a/commit-reach.h b/commit-reach.h
index f41d8f6ba3..acf3b2d188 100644
--- a/commit-reach.h
+++ b/commit-reach.h
@@ -1,9 +1,9 @@
 #ifndef __COMMIT_REACH_H__
 #define __COMMIT_REACH_H__
 
-#include "commit.h"
-#include "commit-slab.h"
+#include "commit-slab-decl.h"
 
+struct commit;
 struct commit_list;
 struct ref_filter;
 struct object_id;
@@ -55,7 +55,7 @@ enum contains_result {
 	CONTAINS_YES
 };
 
-define_commit_slab(contains_cache, enum contains_result);
+define_shared_commit_slab(contains_cache, enum contains_result);
 
 int commit_contains(struct ref_filter *filter, struct commit *commit,
 		    struct commit_list *list, struct contains_cache *cache);
-- 
2.19.0


  reply	other threads:[~2018-09-20 15:35 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-09-19  0:15 [PATCH 9/9] commit-reach.h: add missing declarations (hdr-check) Ramsay Jones
2018-09-19 23:38 ` Derrick Stolee
2018-09-20 15:35   ` Ramsay Jones [this message]
2018-09-21 17:41     ` Derrick Stolee

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=2809251f-6eba-b0ac-68fe-b972809ccff7@ramsayjones.plus.com \
    --to=ramsay@ramsayjones.plus.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=stolee@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).