Fix/Do not use Res pointers in SE

[carpawell]

Based on and blocked by #1418.

Benchmark of the meta.Get:

Comparing just *Res -> Res:

name               old time/op    new time/op    delta
Get/1_objects-8      9.69µs ±13%    9.68µs ± 7%    ~     (p=0.954 n=10+10)
Get/10_objects-8      102µs ±10%     100µs ± 4%    ~     (p=0.393 n=10+10)
Get/100_objects-8    1.00ms ±10%    1.03ms ±13%    ~     (p=0.280 n=10+10)

name               old alloc/op   new alloc/op   delta
Get/1_objects-8      3.22kB ± 0%    3.22kB ± 0%  -0.25%  (p=0.000 n=9+9)
Get/10_objects-8     33.5kB ± 1%    33.4kB ± 1%  -0.38%  (p=0.010 n=9+10)
Get/100_objects-8     336kB ± 0%     335kB ± 0%    ~     (p=0.059 n=9+9)

name               old allocs/op  new allocs/op  delta
Get/1_objects-8        55.0 ± 0%      54.0 ± 0%  -1.82%  (p=0.000 n=9+9)
Get/10_objects-8        576 ± 1%       565 ± 1%  -1.88%  (p=0.001 n=10+10)
Get/100_objects-8     5.80k ± 0%     5.70k ± 0%  -1.71%  (p=0.000 n=9+9)

Comparing Get(Prm) *Res -> Get(Prm, *Res):

name               old time/op    new time/op    delta
Get/1_objects-8      9.69µs ±13%    9.87µs ±21%    ~     (p=0.842 n=10+9)
Get/10_objects-8      102µs ±10%      99µs ± 5%    ~     (p=0.211 n=10+9)
Get/100_objects-8    1.00ms ±10%    1.00ms ± 5%    ~     (p=0.579 n=10+10)

name               old alloc/op   new alloc/op   delta
Get/1_objects-8      3.22kB ± 0%    3.22kB ± 0%  -0.25%  (p=0.000 n=9+9)
Get/10_objects-8     33.5kB ± 1%    33.5kB ± 1%    ~     (p=0.170 n=9+10)
Get/100_objects-8     336kB ± 0%     334kB ± 0%  -0.37%  (p=0.000 n=9+9)

name               old allocs/op  new allocs/op  delta
Get/1_objects-8        55.0 ± 0%      54.0 ± 0%  -1.82%  (p=0.000 n=9+9)
Get/10_objects-8        576 ± 1%       570 ± 0%  -1.02%  (p=0.026 n=10+8)
Get/100_objects-8     5.80k ± 0%     5.69k ± 0%  -1.75%  (p=0.000 n=9+9)

Seems almost equal but I do not like creating Res struct on the caller side and do not see real use cases when that could help much.

[codecov]

Codecov Report

Merging #1460 (bf97cd0) into master (7270829) will decrease coverage by 0.00%.
The diff coverage is 55.45%.

❗ Current head bf97cd0 differs from pull request most recent head 61b309b. Consider uploading reports for the commit 61b309b to get more accurate results

@@            Coverage Diff             @@
##           master    #1460      +/-   ##
==========================================
- Coverage   36.44%   36.44%   -0.01%     
==========================================
  Files         304      304              
  Lines       17847    17843       -4     
==========================================
- Hits         6505     6502       -3     
+ Misses      10790    10789       -1     
  Partials      552      552              

Impacted Files	Coverage Δ
pkg/local_object_storage/blobstor/delete_big.go	0.00% <0.00%> (ø)
pkg/local_object_storage/blobstor/delete_small.go	0.00% <0.00%> (ø)
pkg/local_object_storage/blobstor/get_range_big.go	0.00% <0.00%> (ø)
...g/local_object_storage/blobstor/get_range_small.go	0.00% <0.00%> (ø)
pkg/local_object_storage/engine/container.go	0.00% <0.00%> (ø)
pkg/local_object_storage/engine/delete.go	0.00% <0.00%> (ø)
pkg/local_object_storage/shard/container.go	0.00% <0.00%> (ø)
pkg/local_object_storage/shard/exists.go	0.00% <0.00%> (ø)
pkg/local_object_storage/shard/move.go	0.00% <0.00%> (ø)
pkg/local_object_storage/shard/range.go	0.00% <0.00%> (ø)
... and 36 more

---

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 b432558...61b309b. Read the comment docs.

[cthulhu-rider]

No improvements according to provided results.

[carpawell]

No improvements according to provided results.

What improvements did you expect? We have got rid of one allocation for one Get call. Also, res does not escape to the heap so it also should help GC (I do not know how to test that). My bench is kinda trivial, I may rewrite it to get all the objects for every test (for 1/10/100 objects) then it would have more informative results.

[cthulhu-rider]

@carpawell oh I meant no improvements between 2 provided results in PR desc.

[carpawell]

oh I meant no improvements between 2 provided results in PR desc

Yes, no improvements, so I described why I have decided to choose the first one.

[fyrchik]

@carpawell rebase, please

[carpawell]

@carpawell rebase, please

#1418 must be merged first.

[carpawell]

@carpawell rebase, please

Done, also, reorganized benchmark a little.

[carpawell]

Only updated the changelog.