Tom Lane wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> It would be interesting to see some gprof or oprofile output from that
>> test. I went back and dug up the results that I got when I profiled
>> this patch during initial development, and my version of the patch
>> applied, the profile looked like this on my system:
>
> Were you testing with a temp table? The lack of XLogInsert in your
> profile is striking. Stefan's results upthread had it at the top,
> and I got more or less the same thing here (didn't keep my numbers
> unfortunately).
I guess that profile was for the wal bypass case and it looks fairly
similiar to what I get here(lineitem table into tmpfs - though only 30M
rows this time to keep VM pressure low):
samples % symbol name
286197 17.1997 DoCopy
232958 14.0002 CopyReadLine
99762 5.9954 DecodeNumber
92751 5.5741 heap_fill_tuple
84439 5.0746 pg_verify_mbstr_len
65421 3.9316 InputFunctionCall
62502 3.7562 DecodeDate
53565 3.2191 heap_form_tuple
47731 2.8685 ParseDateTime
41206 2.4764 DecodeDateTime
39936 2.4001 pg_next_dst_boundary
36093 2.1691 AllocSetAlloc
33967 2.0413 heap_compute_data_size
29921 1.7982 float4in
27227 1.6363 DetermineTimeZoneOffset
25622 1.5398 pg_atoi
24703 1.4846 pg_mblen
24495 1.4721 .plt
23912 1.4371 pg_mbstrlen_with_len
23448 1.4092 bpchar_input
20033 1.2039 date2j
16331 0.9815 date_in
15684 0.9426 bpcharin
14819 0.8906 PageAddItem
14261 0.8571 ValidateDate
Stefan
