minor fixups for gs/commit-graph-path-filter
cover minor fixups for gs/commit-graph-path-filter
These are just a few bits I noticed in the test-tool helper when the topic hit next (my -Wunused-parameter patch complained that we never looked at argc).
[1/2]: test-bloom: fix some whitespace issues [2/2]: test-bloom: check that we have expected arguments
t/helper/test-bloom.c | 28 ++++++++++++++++++++-------- 1 file changed, 20 insertions(+), 8 deletions(-)
-Peff
PATCH 1/2 test-bloom: fix some whitespace issues
--- t/helper/test-bloom.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
t/helper/test-bloom.c +5 -5
@@ -27,7 +27,7 @@ static void print_bloom_filter(struct bloom_filter *filter) { }printf ("Filter_Length:%d\n" , (int )filter->len);printf ("Filter_Data:" ); -for (i =0 ; i < filter->len; i++){ +for (i =0 ; i < filter->len; i++) {printf ("%02x|" , filter->data[i]); }printf ("\n" );
@@ -50,13 +50,13 @@ int cmd__bloom(int argc, const char **argv)printf ("Murmur3 Hash with seed=0:0x%08x\n" , hashed); } -if (!strcmp (argv[1 ],"generate_filter" )) { +if (!strcmp (argv[1 ],"generate_filter" )) {struct bloom_filter filter;int i =2 ; filter.len = (settings.bits_per_entry +BITS_PER_WORD -1 ) /BITS_PER_WORD ; filter.data =xcalloc (filter.len,sizeof (unsigned char )); -if (!argv[2 ]){ +if (!argv[2 ]) {die ("at least one input string expected" ); }
@@ -68,7 +68,7 @@ int cmd__bloom(int argc, const char **argv)print_bloom_filter (&filter); } -if (!strcmp (argv[1 ],"get_filter_for_commit" )) { +if (!strcmp (argv[1 ],"get_filter_for_commit" )) {struct object_id oid;const char *end;if (parse_oid_hex (argv[2 ], &oid, &end))
@@ -78,4 +78,4 @@ int cmd__bloom(int argc, const char **argv) }return 0 ; -} +}
Signature · 1 line
2.26.2.827.g3c1233342b
Re: [PATCH 1/2] test-bloom: fix some whitespace issues
Quoted 10 lines · Jeff King
Signed-off-by Jeff King <[email protected]>--- t/helper/test-bloom.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)t/helper/test-bloom.c +0
-0
t/helper/test-bloom.c static void print_bloom_filter(struct bloom_filter *filter) { · line 30
Quoted 5 lines
t/helper/test-bloom.c +1
-1@@ -27,7 +27,7 @@ static void print_bloom_filter(struct bloom_filter *filter) { }printf ("Filter_Length:%d\n" , (int )filter->len);printf ("Filter_Data:" ); -for (i =0 ; i < filter->len; i++){ +for (i =0 ; i < filter->len; i++) {
Thanks for fixing the spacing, but I wonder if these braces should be here at all. Since the body is one line long, maybe this should just be:
for (i = 0; i < filter->len; i++)
printf("%02x|", filter->data[i]);t/helper/test-bloom.c int cmd__bloom(int argc, const char **argv) · line 53
Quoted 9 lines
t/helper/test-bloom.c +1
-1printf ("%02x|" , filter->data[i]); }printf ("\n" );@@ -50,13 +50,13 @@ int cmd__bloom(int argc, const char **argv)printf ("Murmur3 Hash with seed=0:0x%08x\n" , hashed); } -if (!strcmp (argv[1 ],"generate_filter" )) { +if (!strcmp (argv[1 ],"generate_filter" )) {
This spot looks good, and ditto for fixing the indentation.
t/helper/test-bloom.c int cmd__bloom(int argc, const char **argv) · line 59
Quoted 7 lines
t/helper/test-bloom.c +1
-1struct bloom_filter filter;int i =2 ; filter.len = (settings.bits_per_entry +BITS_PER_WORD -1 ) /BITS_PER_WORD ; filter.data =xcalloc (filter.len,sizeof (unsigned char )); -if (!argv[2 ]){ +if (!argv[2 ]) {
Same note here about the brace.
Quoted 20 lines
t/helper/test-bloom.c +2
-3die ("at least one input string expected" ); }@@ -68,7 +68,7 @@ int cmd__bloom(int argc, const char **argv)print_bloom_filter (&filter); } -if (!strcmp (argv[1 ],"get_filter_for_commit" )) { +if (!strcmp (argv[1 ],"get_filter_for_commit" )) {struct object_id oid;const char *end;if (parse_oid_hex (argv[2 ], &oid, &end))@@ -78,4 +78,4 @@ int cmd__bloom(int argc, const char **argv) }return 0 ; -} +} --
Quoted 2 lines
2.26.2.827.g3c1233342b
Thanks,
Taylor
Re: [PATCH 1/2] test-bloom: fix some whitespace issues
Quoted 20 lines · Taylor Blau
Quoted 10 lines · Jeff King
Signed-off-by Jeff King <[email protected]>--- t/helper/test-bloom.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)t/helper/test-bloom.c +0
-0t/helper/test-bloom.c static void print_bloom_filter(struct bloom_filter *filter) { · line 30
Quoted 5 lines
t/helper/test-bloom.c +1
-1@@ -27,7 +27,7 @@ static void print_bloom_filter(struct bloom_filter *filter) { }printf ("Filter_Length:%d\n" , (int )filter->len);printf ("Filter_Data:" ); -for (i =0 ; i < filter->len; i++){ +for (i =0 ; i < filter->len; i++) {Thanks for fixing the spacing, but I wonder if these braces should be here at all. Since the body is one line long, maybe this should just be:
Quoted 2 lines
for (i = 0; i < filter->len; i++)
printf("%02x|", filter->data[i]);I have to admit that I don't care either way, and I think we spend too much time quibbling about braces or not-braces. It was really the bad indentation that I cared about most.
-Peff
PATCH 2/2 test-bloom: check that we have expected arguments
In reply to Jeff KingIf "test-tool bloom" is not fed a command, or if arguments are missing for some commands, it will just segfault. Let's check argc and write a friendlier usage message.
--- t/helper/test-bloom.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-)
t/helper/test-bloom.c +16 -4
@@ -43,10 +43,21 @@ static void get_bloom_filter_for_commit(const struct object_id *commit_oid)print_bloom_filter (filter); } +static const char *bloom_usage ="\n" +" test-tool bloom get_murmer3 <string>\n" +" test-tool bloom generate_filter <string> [<string>...]\n" +" test-tool get_filter_for_commit <commit-hex>\n" ; +int cmd__bloom (int argc,const char **argv) { +if (argc <2 ) +usage (bloom_usage); +if (!strcmp (argv[1 ],"get_murmur3" )) { -uint32_t hashed =murmur3_seeded (0 , argv[2 ],strlen (argv[2 ])); +uint32_t hashed; +if (argc <3 ) +usage (bloom_usage); + hashed =murmur3_seeded (0 , argv[2 ],strlen (argv[2 ]));printf ("Murmur3 Hash with seed=0:0x%08x\n" , hashed); }
@@ -56,9 +67,8 @@ int cmd__bloom(int argc, const char **argv) filter.len = (settings.bits_per_entry +BITS_PER_WORD -1 ) /BITS_PER_WORD ; filter.data =xcalloc (filter.len,sizeof (unsigned char )); -if (!argv[2 ]) { -die ("at least one input string expected" ); - } +if (argc -1 < i) +usage (bloom_usage);while (argv[i]) {add_string_to_filter (argv[i], &filter);
@@ -71,6 +81,8 @@ int cmd__bloom(int argc, const char **argv)if (!strcmp (argv[1 ],"get_filter_for_commit" )) {struct object_id oid;const char *end; +if (argc <3 ) +usage (bloom_usage);if (parse_oid_hex (argv[2 ], &oid, &end))die ("cannot parse oid '%s'" , argv[2 ]);init_bloom_filters ();
Signature · 1 line
2.26.2.827.g3c1233342b
Re: [PATCH 2/2] test-bloom: check that we have expected arguments
Quoted 6 lines · Jeff King
If "test-tool bloom" is not fed a command, or if arguments are missing for some commands, it will just segfault. Let's check argc and write a friendlier usage message.
Signed-off-by Jeff King <[email protected]>---
Looks all good.
Thanks,
Taylor
Quoted 4 lines · Jeff King
These are just a few bits I noticed in the test-tool helper when the topic hit next (my -Wunused-parameter patch complained that we never looked at argc).
Quoted 8 lines
[1/2]: test-bloom: fix some whitespace issues [2/2]: test-bloom: check that we have expected arguments
t/helper/test-bloom.c | 28 ++++++++++++++++++++-------- 1 file changed, 20 insertions(+), 8 deletions(-)-Peff
Thank you for doing this!
Both patches look good to me.
I also don't care about the brace/no-brace thing that Taylor brought up for 1/2.
Cheers!
Garima Singh
Quoted 12 lines · Garima Singh
Quoted 10 lines · Jeff King
These are just a few bits I noticed in the test-tool helper when the topic hit next (my -Wunused-parameter patch complained that we never looked at argc).
[1/2]: test-bloom: fix some whitespace issues [2/2]: test-bloom: check that we have expected arguments
t/helper/test-bloom.c | 28 ++++++++++++++++++++-------- 1 file changed, 20 insertions(+), 8 deletions(-)
Quoted 7 lines
Quoted 2 lines
-Peff
Thank you for doing this!
Both patches look good to me.
I also don't care about the brace/no-brace thing that Taylor brought up for 1/2.
To be clear, I don't care about them either ;). Maybe it's time that we relax that rule (if it seems that a good number of us don't mind it either way)..?
Quoted 2 lines
Cheers!
Garima Singh
Thanks,
Taylor
Quoted 3 lines · Taylor Blau
To be clear, I don't care about them either ;). Maybe it's time that we relax that rule (if it seems that a good number of us don't mind it either way)..?
I'd prefer our reviewers (even more preferably the authors) to be careful about new code, but it probably is not worth reviewing patches that only fix them and do nothing else.
Quoted 3 lines · Jeff King
These are just a few bits I noticed in the test-tool helper when the topic hit next (my -Wunused-parameter patch complained that we never looked at argc).
I think I'll add this one to those few bits.
Old version of this change was sent here:
<[email protected]>
But that version doesn't have the fixup for sh script.
Garima Singh: Could you please change your editor to add final new line?
I've take another look into bloom.h.
I think we should drop BITS_PER_WORD definition and use CHAR_BIT instead. It's a standard definition.
To me, a WORD is an int, at least I was told that when I was still in university and study about computer science.
-----------------8<---------------------
From: =?UTF-8?q?=C4=90o=C3=A0n=20Tr=E1=BA=A7n=20C=C3=B4ng=20Danh?=
<[email protected]>
Date: Thu, 23 Apr 2020 20:24:50 +0700
Subject: [PATCH] bloom: fix make sparse warning
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
* We need a final_new_line to make our source code as text file, per POSIX and C specification.
* bloom_filters should be limited to interal linkage only
---
Feel free to fix up to current series
bloom.c | 4 ++-- bloom.h | 2 +- t/helper/test-bloom.c | 2 +- t/t0095-bloom.sh | 2 +- t/t4216-log-bloom.sh | 2 +- 5 files changed, 6 insertions(+), 6 deletions(-)
bloom.c +2 -2
@@ -9,7 +9,7 @@ define_commit_slab(bloom_filter_slab, struct bloom_filter); -struct bloom_filter_slab bloom_filters; +static struct bloom_filter_slab bloom_filters;struct pathmap_hash_entry {struct hashmap_entry entry;
@@ -273,4 +273,4 @@ int bloom_filter_contains(const struct bloom_filter *filter, }return 1 ; -} +}
bloom.h +1 -1
@@ -87,4 +87,4 @@ int bloom_filter_contains(const struct bloom_filter *filter,const struct bloom_key *key, const struct bloom_filter_settings *settings); -#endif +#endif
t/helper/test-bloom.c +1 -1
@@ -3,7 +3,7 @@#include "test-tool.h" #include "commit.h" -struct bloom_filter_settings settings =DEFAULT_BLOOM_FILTER_SETTINGS ; +static struct bloom_filter_settings settings =DEFAULT_BLOOM_FILTER_SETTINGS ;static void add_string_to_filter (const char *data,struct bloom_filter *filter) {struct bloom_key key;
t/t0095-bloom.sh +1 -1
@@ -114,4 +114,4 @@ test_expect_success EXPENSIVE 'get bloom filter for commit with 513 changes' ' test_cmp expect actual ' -test_done +test_done
t/t4216-log-bloom.sh +1 -1
@@ -152,4 +152,4 @@ test_expect_success 'Use Bloom filters if they exist in the latest but not all c test_bloom_filters_used_when_some_filters_are_missing "-- A/B" ' -test_done +test_done
Signature · 5 lines
2.26.2.384.g435bf60bd5 -- Danh
Quoted 6 lines · Danh Doan
Quoted 3 lines · Jeff King
These are just a few bits I noticed in the test-tool helper when the topic hit next (my -Wunused-parameter patch complained that we never looked at argc).
I think I'll add this one to those few bits.
Yeah, they all look sensible (I should have looked for more "No newline" cases.
Quoted 7 lines
I've take another look into
bloom.h.I think we should drop
BITS_PER_WORDdefinition and useCHAR_BITinstead. It's a standard definition.To me, a WORD is an
int, at least I was told that when I was still in university and study about computer science.
Yes, I agree it would be more clear as just CHAR_BIT if we are using single-char words. But I suspect the code could be looking at the bit patterns using larger word sizes (e.g., all of the ewah code uses 64-bit words). That might be worth exploring.
-Peff