Re: gsoc, oprrest function for text search take 2

Heikki Linnakangas wrote:
> Jan Urbański wrote:
>> 26763     3.5451  AllocSetCheck
>
> Make sure you disable assertions before profiling.

Awww, darn. OK, here goes another set of results, without casserts this
time.

=== CVS HEAD ===

number of clients: 10
number of transactions per client: 100000
number of transactions actually processed: 1000000/1000000
tps = 6437.286494 (including connections establishing)
tps = 6438.168927 (excluding connections establishing)

samples  %        symbol name
220443   11.6613  AllocSetAlloc
79355     4.1978  base_yyparse
77230     4.0854  SearchCatCache
56011     2.9629  hash_search_with_hash_value
45946     2.4305  MemoryContextAllocZeroAligned
38577     2.0407  hash_any
36414     1.9263  MemoryContextAlloc
33060     1.7489  AllocSetFree
27218     1.4398  ScanKeywordLookup
25793     1.3644  base_yylex
20579     1.0886  hash_uint32
18867     0.9981  hash_seq_search
18293     0.9677  expression_tree_walker
17696     0.9361  copyObject
16979     0.8982  LockAcquire
14292     0.7560  MemoryContextAllocZero
13117     0.6939  SearchSysCache

=== ts_sel ====

number of clients: 10
number of transactions per client: 100000
number of transactions actually processed: 1000000/1000000
tps = 3216.753677 (including connections establishing)
tps = 3216.996592 (excluding connections establishing)

942096   10.9130  internal_text_pattern_compare
809195    9.3735  bttext_pattern_cmp
659545    7.6400  pg_detoast_datum_packed
628114    7.2759  pg_qsort
603998    6.9966  AllocSetAlloc
581880    6.7403  pglz_decompress
467708    5.4178  DirectFunctionCall2
385854    4.4696  compare_two_textfreqs
160578    1.8601  AllocSetFree
128642    1.4902  swapfunc
112885    1.3076  MemoryContextAlloc
103388    1.1976  SearchCatCache
100387    1.1629  text_to_cstring
99004     1.1468  hash_search_with_hash_value
98444     1.1403  .plt
92664     1.0734  base_yyparse
88511     1.0253  errstart

Not good... Shall I try sorting pg_statistics arrays on text values
instead of frequencies?
BTW: I just noticed some text_to_cstring calls, they came from
elog(DEBUG1)s that I have in my code. But they couldn't have skewn the
results much, could they?

Cheers,
Jan

--
Jan Urbanski
GPG key ID: E583D7D2

ouden estin