btrfs-progs: GCC9 fixes
cover btrfs-progs: GCC9 fixes
This patchset will address the remaining warning when compiling btrfs-progs devel branch with GCC9.
It's based on the following commit:
commit 9a1d86a9ac7384b332db498822585a2255f7d3e6 (david/devel)
Author: David Sterba <[email protected]> Date: Thu Jun 13 20:45:49 2019 +0200
btrfs-progs: build: disable -Waddress-of-packed-member by default
Please note that the 2nd patch mostly replace commit 691656abdc9a ("btrfs-progs: fix gcc9 warning and potentially unaligned access to dev stats") by backporing kernel btrfs_dev_stats_value() to btrfs-progs.
Thus the original fix can be removed.
The remaining warning is -Warray-boundary, but that one looks pretty strange. The code looks good, and I backported an easy version:
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <stddef.h>
struct test_struct {
long long off_0_7;
int offset_8_11;
unsigned char offset_12_12;
} __attribute__ ((packed));
void reset_values(struct test_struct *ptr)
{
memset(&ptr->offset_8_11, 0, sizeof(struct test_struct) - offsetof(struct test_struct, offset_8_11));
}
int main()
{
struct test_struct my_struct = { 0xffff, 0xff, 0xff};
printf("struct=0x%llx start=0x%llx len=0x%x\n",
&my_struct, &my_struct.offset_8_11, sizeof(struct test_struct) - offsetof(struct test_struct, offset_8_11));
reset_values(&my_struct);
printf("0x%lx 0x%x 0x%x\n", my_struct.off_0_7, my_struct.offset_8_11, my_struct.offset_12_12);
return 0;
}Which doesn't reproduce the warning.
Thus looks like a false warning and a bug in gcc.
Qu Wenruo (4):
btrfs-progs: constify extent buffer reader
btrfs-progs: Fix -Waddress-of-packed-member warning in
btrfs_dev_stats_values callers
btrfs-progs: Remove unnecessary fallthrough attribute in
test_num_disk_vs_raid()
btrfs-progs: Fix Wformat-overflow warning in cmds-receive.c
cmds-receive.c | 4 ++-- ctree.h | 13 +++++++++++++ extent_io.c | 4 ++-- extent_io.h | 4 ++-- print-tree.c | 21 ++++++--------------- utils.c | 1 - 6 files changed, 25 insertions(+), 22 deletions(-)
Signature · 1 line
2.22.0
PATCH 1/4 btrfs-progs: constify extent buffer reader
Add const prefix for the following parameters:
- @eb of memcmp_extent_buffer()
- @eb of read_extent_buffer()
This backports kernel commit 1cbb1f454e53 ("btrfs: struct-funcs, constify readers") to btrfs-progs.
--- extent_io.c | 4 ++-- extent_io.h | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-)
extent_io.c +2 -2
@@ -963,13 +963,13 @@ int clear_extent_buffer_dirty(struct extent_buffer *eb)return 0 ; } -int memcmp_extent_buffer (struct extent_buffer *eb,const void *ptrv, +int memcmp_extent_buffer (const struct extent_buffer *eb,const void *ptrv,unsigned long start,unsigned long len) {return memcmp (eb->data + start, ptrv, len); } -void read_extent_buffer (struct extent_buffer *eb,void *dst, +void read_extent_buffer (const struct extent_buffer *eb,void *dst,unsigned long start,unsigned long len) {memcpy (dst, eb->data + start, len);
extent_io.h +2 -2
@@ -154,9 +154,9 @@ void free_extent_buffer_nocache(struct extent_buffer *eb);int read_extent_from_disk (struct extent_buffer *eb,unsigned long offset,unsigned long len);int write_extent_to_disk (struct extent_buffer *eb); -int memcmp_extent_buffer (struct extent_buffer *eb,const void *ptrv, +int memcmp_extent_buffer (const struct extent_buffer *eb,const void *ptrv,unsigned long start,unsigned long len); -void read_extent_buffer (struct extent_buffer *eb,void *dst, +void read_extent_buffer (const struct extent_buffer *eb,void *dst,unsigned long start,unsigned long len);void write_extent_buffer (struct extent_buffer *eb,const void *src,unsigned long start,unsigned long len);
Signature · 1 line
2.22.0
PATCH 2/4 btrfs-progs: Fix -Waddress-of-packed-member warning in btrfs_dev_stats_values callers
In reply to Qu Wenruo[BUG]
GCC 9.1.0 will report the following error when compiling btrfs-progs:
In file included from print-tree.c:24:
ctree.h: In function 'btrfs_dev_stats_values':
ctree.h:2408:9: warning: taking address of packed member of 'struct btrfs_dev_stats_item' may result in an unaligned pointer value [-Waddress-of-packed-member]
2408 | return p->values;
| ^
[FIX]
Follow the kernel way of accessing dev stats by using btrfs_dev_stats_value(eb, ptr, index).
So that we don't need to bother accessing the packed member.
This also unifies the helper function in kernel and btrfs-progs.
--- ctree.h | 13 +++++++++++++ print-tree.c | 21 ++++++--------------- 2 files changed, 19 insertions(+), 15 deletions(-)
ctree.h +13 -0
@@ -2395,6 +2395,19 @@ static inline struct btrfs_disk_balance_args* btrfs_balance_item_sys(return &p->sys; } +static inline u64 btrfs_dev_stats_value (const struct extent_buffer *eb, +const struct btrfs_dev_stats_item *ptr, +int index) +{ +u64 val; + +read_extent_buffer (eb, &val, + offsetof(struct btrfs_dev_stats_item , values) + + ((unsigned long )ptr) + (index *sizeof (u64)), +sizeof (val)); +return val; +} +/* * this returns the number of bytes used by the item on disk, minus the * size of any extent headers. If a file is compressed on disk, this is
print-tree.c +6 -15
@@ -953,32 +953,23 @@ static void print_balance_item(struct extent_buffer *eb,static void print_dev_stats (struct extent_buffer *eb,struct btrfs_dev_stats_item *stats,u32 size) { -struct btrfs_dev_stats_item *item; -const unsigned long offset = (unsigned long )stats;u32 known =BTRFS_DEV_STAT_VALUES_MAX *sizeof (__le64);int i; - item = (struct btrfs_dev_stats_item *)(eb->data + offset); -printf ("\t\tdevice stats\n" );printf ("\t\twrite_errs %llu read_errs %llu flush_errs %llu corruption_errs %llu generation %llu\n" , - (unsigned long long )get_unaligned_le64 ( - &item->values[BTRFS_DEV_STAT_WRITE_ERRS ]), - (unsigned long long )get_unaligned_le64 ( - &item->values[BTRFS_DEV_STAT_READ_ERRS ]), - (unsigned long long )get_unaligned_le64 ( - &item->values[BTRFS_DEV_STAT_FLUSH_ERRS ]), - (unsigned long long )get_unaligned_le64 ( - &item->values[BTRFS_DEV_STAT_CORRUPTION_ERRS ]), - (unsigned long long )get_unaligned_le64 ( - &item->values[BTRFS_DEV_STAT_GENERATION_ERRS ])); +btrfs_dev_stats_value (eb, stats,BTRFS_DEV_STAT_WRITE_ERRS ), +btrfs_dev_stats_value (eb, stats,BTRFS_DEV_STAT_READ_ERRS ), +btrfs_dev_stats_value (eb, stats,BTRFS_DEV_STAT_FLUSH_ERRS ), +btrfs_dev_stats_value (eb, stats,BTRFS_DEV_STAT_CORRUPTION_ERRS ), +btrfs_dev_stats_value (eb, stats,BTRFS_DEV_STAT_GENERATION_ERRS ));if (known < size) {printf ("\t\tunknown stats item bytes %u" , size - known);for (i =BTRFS_DEV_STAT_VALUES_MAX ; i *sizeof (__le64) < size; i++) {printf ("\t\tunknown item %u offset %zu value %llu\n" , i, i *sizeof (__le64), - (unsigned long long )le64_to_cpu (&item->values[i])); +btrfs_dev_stats_value (eb, stats, i)); } } }
Signature · 1 line
2.22.0
PATCH 3/4 btrfs-progs: Remove unnecessary fallthrough attribute in test_num_disk_vs_raid()
In reply to Qu Wenruo--- utils.c | 1 - 1 file changed, 1 deletion(-)
utils.c +0 -1
@@ -1928,7 +1928,6 @@ int test_num_disk_vs_raid(u64 metadata_profile, u64 data_profile, __attribute__ ((fallthrough));case 1 : allowed |=BTRFS_BLOCK_GROUP_DUP ; - __attribute__ ((fallthrough)); }if (dev_cnt >1 && profile &BTRFS_BLOCK_GROUP_DUP ) {
Signature · 1 line
2.22.0
Re: [PATCH 3/4] btrfs-progs: Remove unnecessary fallthrough attribute in test_num_disk_vs_raid()
Even though this is a simple fix, the warning should be here and the versio that triggered it.
Quoted 19 lines · Qu Wenruo
Signed-off-by Qu Wenruo <[email protected]>--- utils.c | 1 - 1 file changed, 1 deletion(-)utils.c +0
-1@@ -1928,7 +1928,6 @@ int test_num_disk_vs_raid(u64 metadata_profile, u64 data_profile, __attribute__ ((fallthrough));case 1 : allowed |=BTRFS_BLOCK_GROUP_DUP ; - __attribute__ ((fallthrough)); }if (dev_cnt >1 && profile &BTRFS_BLOCK_GROUP_DUP ) {Signature · 1 line
2.22.0
PATCH 4/4 btrfs-progs: Fix Wformat-overflow warning in cmds-receive.c
In reply to Qu Wenruo[BUG]
When compiling btrfs-progs with GCC 9 (9.1.0), we got the following warnings:
In file included from utils.h:30,
from cmds-receive.c:45:
cmds-receive.c: In function 'process_subvol':
messages.h:42:3: warning: '%s' directive argument is null [-Wformat-overflow=]
42 | __btrfs_error((fmt), ##__VA_ARGS__); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cmds-receive.c:178:3: note: in expansion of macro 'error'
178 | error("subvol: another one already started, path buf: %s",
| ^~~~~
[CC] cmds-inspect-tree-stats.o
cmds-receive.c: In function 'process_snapshot':
messages.h:42:3: warning: '%s' directive argument is null [-Wformat-overflow=]
42 | __btrfs_error((fmt), ##__VA_ARGS__); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cmds-receive.c:248:3: note: in expansion of macro 'error'
248 | error("snapshot: another one already started, path buf: %s",
| ^~~~~[FIX]
We're using wrong member for the error output.
Fix the member to output.
--- cmds-receive.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
cmds-receive.c +2 -2
@@ -176,7 +176,7 @@ static int process_subvol(const char *path, const u8 *uuid, u64 ctransid, }if (rctx->cur_subvol_path[0 ]) {error ("subvol: another one already started, path buf: %s" , - rctx->cur_subvol.path); + rctx->cur_subvol_path); ret = -EINVAL ; goto out; }
@@ -246,7 +246,7 @@ static int process_snapshot(const char *path, const u8 *uuid, u64 ctransid, }if (rctx->cur_subvol_path[0 ]) {error ("snapshot: another one already started, path buf: %s" , - rctx->cur_subvol.path); + rctx->cur_subvol_path); ret = -EINVAL ; goto out; }
Signature · 1 line
2.22.0
Quoted 8 lines · Qu Wenruo
This patchset will address the remaining warning when compiling btrfs-progs devel branch with GCC9.
It's based on the following commit:
commit9a1d86a9ac7384b332db498822585a2255f7d3e6(david/devel)Author: David Sterba <[email protected]> Date: Thu Jun 13 20:45:49 2019 +0200
Quoted 7 lines
btrfs-progs: build: disable -Waddress-of-packed-member by default
Please note that the 2nd patch mostly replace commit
691656abdc9a("btrfs-progs: fix gcc9 warning and potentially unaligned access to dev stats") by backporing kernelbtrfs_dev_stats_value()to btrfs-progs.
Thus the original fix can be removed.
Yeah, the backport is better, patch replaced.
Quoted 35 lines
The remaining warning is -Warray-boundary, but that one looks pretty strange. The code looks good, and I backported an easy version:
#include <stdio.h> #include <stdlib.h> #include <string.h> #include <stddef.h> struct test_struct { long long off_0_7; int offset_8_11; unsigned char offset_12_12; } __attribute__ ((packed)); void reset_values(struct test_struct *ptr) { memset(&ptr->offset_8_11, 0, sizeof(struct test_struct) - offsetof(struct test_struct, offset_8_11)); } int main() { struct test_struct my_struct = { 0xffff, 0xff, 0xff}; printf("struct=0x%llx start=0x%llx len=0x%x\n", &my_struct, &my_struct.offset_8_11, sizeof(struct test_struct) - offsetof(struct test_struct, offset_8_11)); reset_values(&my_struct); printf("0x%lx 0x%x 0x%x\n", my_struct.off_0_7, my_struct.offset_8_11, my_struct.offset_12_12); return 0; }Which doesn't reproduce the warning.
Thus looks like a false warning and a bug in gcc.
Quoted 7 lines
Qu Wenruo (4):
btrfs-progs: constify extent buffer reader
btrfs-progs: Fix -Waddress-of-packed-member warning in
btrfs_dev_stats_valuescallers
btrfs-progs: Remove unnecessary fallthrough attribute intest_num_disk_vs_raid()btrfs-progs: Fix Wformat-overflow warning in
cmds-receive.c
1-4 applied, thanks.