cachepc-linux

Fork of AMDESE/linux with modifications for CachePC side-channel attack
git clone https://git.sinitax.com/sinitax/cachepc-linux
Log | Files | Refs | README | LICENSE | sfeed.txt

callchain-overhead-calculation.txt (3430B)


      1Overhead calculation
      2--------------------
      3The overhead can be shown in two columns as 'Children' and 'Self' when
      4perf collects callchains.  The 'self' overhead is simply calculated by
      5adding all period values of the entry - usually a function (symbol).
      6This is the value that perf shows traditionally and sum of all the
      7'self' overhead values should be 100%.
      8
      9The 'children' overhead is calculated by adding all period values of
     10the child functions so that it can show the total overhead of the
     11higher level functions even if they don't directly execute much.
     12'Children' here means functions that are called from another (parent)
     13function.
     14
     15It might be confusing that the sum of all the 'children' overhead
     16values exceeds 100% since each of them is already an accumulation of
     17'self' overhead of its child functions.  But with this enabled, users
     18can find which function has the most overhead even if samples are
     19spread over the children.
     20
     21Consider the following example; there are three functions like below.
     22
     23-----------------------
     24void foo(void) {
     25    /* do something */
     26}
     27
     28void bar(void) {
     29    /* do something */
     30    foo();
     31}
     32
     33int main(void) {
     34    bar()
     35    return 0;
     36}
     37-----------------------
     38
     39In this case 'foo' is a child of 'bar', and 'bar' is an immediate
     40child of 'main' so 'foo' also is a child of 'main'.  In other words,
     41'main' is a parent of 'foo' and 'bar', and 'bar' is a parent of 'foo'.
     42
     43Suppose all samples are recorded in 'foo' and 'bar' only.  When it's
     44recorded with callchains the output will show something like below
     45in the usual (self-overhead-only) output of perf report:
     46
     47----------------------------------
     48Overhead  Symbol
     49........  .....................
     50  60.00%  foo
     51          |
     52          --- foo
     53              bar
     54              main
     55              __libc_start_main
     56
     57  40.00%  bar
     58          |
     59          --- bar
     60              main
     61              __libc_start_main
     62----------------------------------
     63
     64When the --children option is enabled, the 'self' overhead values of
     65child functions (i.e. 'foo' and 'bar') are added to the parents to
     66calculate the 'children' overhead.  In this case the report could be
     67displayed as:
     68
     69-------------------------------------------
     70Children      Self  Symbol
     71........  ........  ....................
     72 100.00%     0.00%  __libc_start_main
     73          |
     74          --- __libc_start_main
     75
     76 100.00%     0.00%  main
     77          |
     78          --- main
     79              __libc_start_main
     80
     81 100.00%    40.00%  bar
     82          |
     83          --- bar
     84              main
     85              __libc_start_main
     86
     87  60.00%    60.00%  foo
     88          |
     89          --- foo
     90              bar
     91              main
     92              __libc_start_main
     93-------------------------------------------
     94
     95In the above output, the 'self' overhead of 'foo' (60%) was add to the
     96'children' overhead of 'bar', 'main' and '\_\_libc_start_main'.
     97Likewise, the 'self' overhead of 'bar' (40%) was added to the
     98'children' overhead of 'main' and '\_\_libc_start_main'.
     99
    100So '\_\_libc_start_main' and 'main' are shown first since they have
    101same (100%) 'children' overhead (even though they have zero 'self'
    102overhead) and they are the parents of 'foo' and 'bar'.
    103
    104Since v3.16 the 'children' overhead is shown by default and the output
    105is sorted by its values. The 'children' overhead is disabled by
    106specifying --no-children option on the command line or by adding
    107'report.children = false' or 'top.children = false' in the perf config
    108file.