Fix bp_embedded_type enum definition

[loli10K]

Motivation and Context

With the addition of BP_EMBEDDED_TYPE_REDACTED in 30af21b a couple of codepaths make wrong assumptions and could potentially result in errors.

(gdb) ptype bp_embedded_type_t
type = enum bp_embedded_type {BP_EMBEDDED_TYPE_DATA, BP_EMBEDDED_TYPE_RESERVED, BP_EMBEDDED_TYPE_REDACTED, NUM_BP_EMBEDDED_TYPES = 1}
(gdb) p/d BP_EMBEDDED_TYPE_REDACTED
$1 = 2
(gdb) 

With a bp (etype = BP_EMBEDDED_TYPE_REDACTED) we could panic in zfs_blkptr_verify():
https://github.com/zfsonlinux/zfs/blob/30af21b02569ac192f52ce6e6511015f8a8d5729/module/zfs/zio.c#L910-L915

Also we would not be able to receive records with drr_etype = BP_EMBEDDED_TYPE_REDACTED:
https://github.com/zfsonlinux/zfs/blob/30af21b02569ac192f52ce6e6511015f8a8d5729/module/zfs/dmu_recv.c#L1849-L1850

And finally we will fail an assetion in dmu_write_embedded():
https://github.com/zfsonlinux/zfs/blob/30af21b02569ac192f52ce6e6511015f8a8d5729/module/zfs/dmu.c#L1270-L1276

I'm not sure if BP_EMBEDDED_TYPE_RESERVED was supposed to be BP_EMBEDDED_TYPE_REDACTED all along. I could not find any usage of BP_SET_REDACTED(), dmu_buf_redact() and dmu_redact() in the code.

Description

Simply update NUM_BP_EMBEDDED_TYPES in bp_embedded_type_t

How Has This Been Tested?

Not tested, pushed as draft.

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Performance enhancement (non-breaking change which improves efficiency)
• Code cleanup (non-breaking change which makes code smaller or more readable)
• Breaking change (fix or feature that would cause existing functionality to change)
• Documentation (a change to man pages or other documentation)

Checklist:

• My code follows the ZFS on Linux code style requirements.
• I have updated the documentation accordingly.
• I have read the contributing document.
• I have added tests to cover my changes.
• All new and existing tests passed.
• All commit messages are properly formatted and contain Signed-off-by.

[codecov]

Codecov Report

Merging #8951 into master will increase coverage by 0.09%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #8951      +/-   ##
==========================================
+ Coverage   78.61%    78.7%   +0.09%     
==========================================
  Files         388      388              
  Lines      120071   120071              
==========================================
+ Hits        94388    94503     +115     
+ Misses      25683    25568     -115

Flag	Coverage Δ
#kernel	79.46% <100%> (+0.01%)	⬆️
#user	66.34% <100%> (+0.31%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 8f12a4f...66caab6. Read the comment docs.

[chrisrd]

How about fixing struct zdb_cb in cmd/zdb/zdb.c as well? E.g.:

typedef struct zdb_cb {
        ...
        uint64_t        zcb_embedded_blocks[NUM_BP_EMBEDDED_TYPES - 1];
        uint64_t        zcb_embedded_histogram[NUM_BP_EMBEDDED_TYPES - 1]
            [BPE_PAYLOAD_SIZE + 1];
        ...
} zdb_cb_t;

[pcd1193182]

Currently, redaction BPs aren't actually used by redacted receive for performance reasons (need an efficient implementation of redacting large ranges); re-enabling them is on my roadmap.

[behlendorf]

@loli10K thanks for catching this. @pcd1193182 helpfully referred me to the DxOS source where BP_EMBEDDED_TYPE_RESERVED is defined and used for an entirely different feature. While you're here it would be a nice to update that comment. Maybe:

 	BP_EMBEDDED_TYPE_RESERVED, /* Reserved for Delphix byteswap feature. */