[PATCH] btrfs-progs: add extent-tree-v2 support to dump-super

[PATCH] btrfs-progs: add extent-tree-v2 support to dump-super

[Anand Jain]

Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
 kernel-shared/print-tree.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/kernel-shared/print-tree.c b/kernel-shared/print-tree.c
index 5c3d14298b58..6b5fd37ab2bc 100644
--- a/kernel-shared/print-tree.c
+++ b/kernel-shared/print-tree.c
@@ -1689,6 +1689,9 @@ static struct readable_flag_entry incompat_flags_array[] = {
 	DEF_INCOMPAT_FLAG_ENTRY(METADATA_UUID),
 	DEF_INCOMPAT_FLAG_ENTRY(RAID1C34),
 	DEF_INCOMPAT_FLAG_ENTRY(ZONED),
+#if EXPERIMENTAL
+	DEF_INCOMPAT_FLAG_ENTRY(EXTENT_TREE_V2),
+#endif
 };
 static const int incompat_flags_num = sizeof(incompat_flags_array) /
 				      sizeof(struct readable_flag_entry);
-- 
2.31.1

Re: [PATCH] btrfs-progs: add extent-tree-v2 support to dump-super

[Qu Wenruo]

On 2022/10/8 17:02, Anand Jain wrote:
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> ---
>   kernel-shared/print-tree.c | 3 +++
>   1 file changed, 3 insertions(+)
>
> diff --git a/kernel-shared/print-tree.c b/kernel-shared/print-tree.c
> index 5c3d14298b58..6b5fd37ab2bc 100644
> --- a/kernel-shared/print-tree.c
> +++ b/kernel-shared/print-tree.c
> @@ -1689,6 +1689,9 @@ static struct readable_flag_entry incompat_flags_array[] = {
>   	DEF_INCOMPAT_FLAG_ENTRY(METADATA_UUID),
>   	DEF_INCOMPAT_FLAG_ENTRY(RAID1C34),
>   	DEF_INCOMPAT_FLAG_ENTRY(ZONED),
> +#if EXPERIMENTAL
> +	DEF_INCOMPAT_FLAG_ENTRY(EXTENT_TREE_V2),
> +#endif

That's in fact one solution I want to go.

But later I found that, we can enhance __print_readable_flag() to
iterate the incompat_flags_array[] with extra check on @supported_flags,
so we can skip the EXPERIMENTAL macro inside the array.

By that, we can reduce the number of EXPERIMENTAL macros, which I
believe is already causing burdens for testing.

Thanks,
Qu
>   };
>   static const int incompat_flags_num = sizeof(incompat_flags_array) /
>   				      sizeof(struct readable_flag_entry);

Re: [PATCH] btrfs-progs: add extent-tree-v2 support to dump-super

[Anand Jain]

On 10/9/22 07:13, Qu Wenruo wrote:
> 
> 
> On 2022/10/8 17:02, Anand Jain wrote:
>> Signed-off-by: Anand Jain <anand.jain@oracle.com>
>> ---
>>   kernel-shared/print-tree.c | 3 +++
>>   1 file changed, 3 insertions(+)
>>
>> diff --git a/kernel-shared/print-tree.c b/kernel-shared/print-tree.c
>> index 5c3d14298b58..6b5fd37ab2bc 100644
>> --- a/kernel-shared/print-tree.c
>> +++ b/kernel-shared/print-tree.c
>> @@ -1689,6 +1689,9 @@ static struct readable_flag_entry 
>> incompat_flags_array[] = {
>>       DEF_INCOMPAT_FLAG_ENTRY(METADATA_UUID),
>>       DEF_INCOMPAT_FLAG_ENTRY(RAID1C34),
>>       DEF_INCOMPAT_FLAG_ENTRY(ZONED),
>> +#if EXPERIMENTAL
>> +    DEF_INCOMPAT_FLAG_ENTRY(EXTENT_TREE_V2),
>> +#endif
> 
> That's in fact one solution I want to go.
> 
> But later I found that, we can enhance __print_readable_flag() to
> iterate the incompat_flags_array[] with extra check on @supported_flags,
> so we can skip the EXPERIMENTAL macro inside the array.
> 
> By that, we can reduce the number of EXPERIMENTAL macros, which I
> believe is already causing burdens for testing.

Yeah, possible. However, it is a cleanup; I wouldn't mix that in the bug 
fix, here.

Thanks, Anand

> 
> Thanks,
> Qu
>>   };
>>   static const int incompat_flags_num = sizeof(incompat_flags_array) /
>>                         sizeof(struct readable_flag_entry);

Re: [PATCH] btrfs-progs: add extent-tree-v2 support to dump-super

[David Sterba]

On Mon, Oct 10, 2022 at 03:43:51PM +0800, Anand Jain wrote:
> On 10/9/22 07:13, Qu Wenruo wrote:
> > On 2022/10/8 17:02, Anand Jain wrote:
> >> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> >> ---
> >>   kernel-shared/print-tree.c | 3 +++
> >>   1 file changed, 3 insertions(+)
> >>
> >> diff --git a/kernel-shared/print-tree.c b/kernel-shared/print-tree.c
> >> index 5c3d14298b58..6b5fd37ab2bc 100644
> >> --- a/kernel-shared/print-tree.c
> >> +++ b/kernel-shared/print-tree.c
> >> @@ -1689,6 +1689,9 @@ static struct readable_flag_entry 
> >> incompat_flags_array[] = {
> >>       DEF_INCOMPAT_FLAG_ENTRY(METADATA_UUID),
> >>       DEF_INCOMPAT_FLAG_ENTRY(RAID1C34),
> >>       DEF_INCOMPAT_FLAG_ENTRY(ZONED),
> >> +#if EXPERIMENTAL
> >> +    DEF_INCOMPAT_FLAG_ENTRY(EXTENT_TREE_V2),
> >> +#endif
> > 
> > That's in fact one solution I want to go.
> > 
> > But later I found that, we can enhance __print_readable_flag() to
> > iterate the incompat_flags_array[] with extra check on @supported_flags,
> > so we can skip the EXPERIMENTAL macro inside the array.
> > 
> > By that, we can reduce the number of EXPERIMENTAL macros, which I
> > believe is already causing burdens for testing.
> 
> Yeah, possible. However, it is a cleanup; I wouldn't mix that in the bug 
> fix, here.

I'll apply your patch as it's a hotfix, what Qu suggests can be done as
consolidation later, though I agree that the number of EXPERIMENTAL
ifdefs should be kept reasonable.