|
Perl is a profligate wastrel when it comes to memory use. There is a saying that to
estimate memory usage of Perl, assume a reasonable algorithm for memory allocation, multiply
that estimate by 10, and while you still may miss the mark, at least you won't be quite so
astonished. This is not absolutely true, but may provide a good grasp of what happens.
Assume that an integer cannot take less than 20 bytes of memory, a float cannot take less
than 24 bytes, a string cannot take less than 32 bytes (all these examples assume 32-bit
architectures, the result are quite a bit worse on 64-bit architectures). If a variable is
accessed in two of three different ways (which require an integer, a float, or a string), the
memory footprint may increase yet another 20 bytes. A sloppy malloc(3) implementation can
inflate these numbers dramatically.
On the opposite end of the scale, a declaration like
may take up to 500 bytes of memory, depending on which release of Perl you're running.
Anecdotal estimates of source-to-compiled code bloat suggest an eightfold increase. This
means that the compiled form of reasonable (normally commented, properly indented etc.) code
will take about eight times more space in memory than the code took on disk.
There are two Perl-specific ways to analyze memory usage: $ENV{PERL_DEBUG_MSTATS} and -DL
command-line switch. The first is available only if Perl is compiled with Perl's malloc(); the
second only if Perl was built with -DDEBUGGING. See the instructions for how to
do this in the INSTALL podpage at the top level of the Perl source tree.
If your perl is using Perl's malloc() and was compiled with the necessary switches (this is
the default), then it will print memory usage statistics after compiling your code when $ENV{PERL_DEBUG_MSTATS}
> 1, and before termination of the program when $ENV{PERL_DEBUG_MSTATS} >=
1. The report format is similar to the following example:
$ PERL_DEBUG_MSTATS=2 perl -e "require Carp"
Memory allocation statistics after compilation: (buckets 4(4)..8188(8192)
14216 free: 130 117 28 7 9 0 2 2 1 0 0
437 61 36 0 5
60924 used: 125 137 161 55 7 8 6 16 2 0 1
74 109 304 84 20
Total sbrk(): 77824/21:119. Odd ends: pad+heads+chain+tail: 0+636+0+2048.
Memory allocation statistics after execution: (buckets 4(4)..8188(8192)
30888 free: 245 78 85 13 6 2 1 3 2 0 1
315 162 39 42 11
175816 used: 265 176 1112 111 26 22 11 27 2 1 1
196 178 1066 798 39
Total sbrk(): 215040/47:145. Odd ends: pad+heads+chain+tail: 0+2192+0+6144.
|
|
It is possible to ask for such a statistic at arbitrary points in your execution using the
mstat() function out of the standard Devel::Peek module.
Here is some explanation of that format:
buckets
SMALLEST(APPROX)..GREATEST(APPROX)
-
Perl's malloc() uses bucketed allocations. Every request is rounded up to the closest
bucket size available, and a bucket is taken from the pool of buckets of that size.
The line above describes the limits of buckets currently in use. Each bucket has two
sizes: memory footprint and the maximal size of user data that can fit into this bucket.
Suppose in the above example that the smallest bucket were size 4. The biggest bucket
would have usable size 8188, and the memory footprint would be 8192.
In a Perl built for debugging, some buckets may have negative usable size. This means
that these buckets cannot (and will not) be used. For larger buckets, the memory footprint
may be one page greater than a power of 2. If so, case the corresponding power of two is
printed in the APPROX field above.
- Free/Used
-
The 1 or 2 rows of numbers following that correspond to the number of buckets of each
size between SMALLEST and GREATEST. In the first row, the sizes
(memory footprints) of buckets are powers of two--or possibly one page greater. In the
second row, if present, the memory footprints of the buckets are between the memory
footprints of two buckets "above".
For example, suppose under the previous example, the memory footprints were
free: 8 16 32 64 128 256 512 1024 2048 4096 8192
4 12 24 48 80
|
|
With non-DEBUGGING perl, the buckets starting from 128 have a
4-byte overhead, and thus an 8192-long bucket may take up to 8188-byte allocations.
Total sbrk(): SBRKed/SBRKs:CONTINUOUS
-
The first two fields give the total amount of memory perl sbrk(2)ed (ess-broken? :-)
and number of sbrk(2)s used. The third number is what perl thinks about continuity of
returned chunks. So long as this number is positive, malloc() will assume that it is
probable that sbrk(2) will provide continuous memory.
Memory allocated by external libraries is not counted.
pad: 0
- The amount of sbrk(2)ed memory needed to keep buckets aligned.
heads: 2192
- Although memory overhead of bigger buckets is kept inside the bucket, for smaller
buckets, it is kept in separate areas. This field gives the total size of these areas.
chain: 0
- malloc() may want to subdivide a bigger bucket into smaller buckets. If only a part of
the deceased bucket is left unsubdivided, the rest is kept as an element of a linked list.
This field gives the total size of these chunks.
tail: 6144
- To minimize the number of sbrk(2)s, malloc() asks for more memory. This field gives the
size of the yet unused part, which is sbrk(2)ed, but never touched.
Below we show how to analyse memory usage by
do 'lib/auto/POSIX/autosplit.ix';
|
|
The file in question contains a header and 146 lines similar to
WARNING: The discussion below supposes 32-bit architecture. In newer releases of
Perl, memory usage of the constructs discussed here is greatly improved, but the story
discussed below is a real-life story. This story is mercilessly terse, and assumes rather more
than cursory knowledge of Perl internals. Type space to continue, `q' to quit. (Actually, you
just want to skip to the next section.)
Here is the itemized list of Perl allocations performed during parsing of this file:
!!! "after" at test.pl line 3.
Id subtot 4 8 12 16 20 24 28 32 36 40 48 56 64 72 80 80+
0 02 13752 . . . . 294 . . . . . . . . . . 4
0 54 5545 . . 8 124 16 . . . 1 1 . . . . . 3
5 05 32 . . . . . . . 1 . . . . . . . .
6 02 7152 . . . . . . . . . . 149 . . . . .
7 02 3600 . . . . . 150 . . . . . . . . . .
7 03 64 . -1 . 1 . . 2 . . . . . . . . .
7 04 7056 . . . . . . . . . . . . . . . 7
7 17 38404 . . . . . . . 1 . . 442 149 . . 147 .
9 03 2078 17 249 32 . . . . 2 . . . . . . . .
|
|
To see this list, insert two warn('!...') statements around the call:
warn('!');
do 'lib/auto/POSIX/autosplit.ix';
warn('!!! "after"');
|
|
and run it with Perl's -DL option. The first warn() will print memory allocation
info before parsing the file and will memorize the statistics at this point (we ignore what it
prints). The second warn() prints increments with respect to these memorized data. This is the
printout shown above.
Different Ids on the left correspond to different subsystems of the perl
interpreter. They are just the first argument given to the perl memory allocation API named
New(). To find what 9 03 means, just grep the perl source for 903.
You'll find it in util.c, function savepvn(). (I know, you wonder why we told you to grep
and then gave away the answer. That's because grepping the source is good for the soul.) This
function is used to store a copy of an existing chunk of memory. Using a C debugger, one can
see that the function was called either directly from gv_init() or via sv_magic(), and that
gv_init() is called from gv_fetchpv()--which was itself called from newSUB(). Please stop to
catch your breath now.
NOTE: To reach this point in the debugger and skip the calls to savepvn() during the
compilation of the main program, you should set a C breakpoint in Perl_warn(), continue until
this point is reached, and then set a C breakpoint in Perl_savepvn(). Note that you may
need to skip a handful of Perl_savepvn() calls that do not correspond to mass production of
CVs (there are more 903 allocations than 146 similar lines of lib/auto/POSIX/autosplit.ix).
Note also that Perl_ prefixes are added by macroization code in perl header files
to avoid conflicts with external libraries.
Anyway, we see that 903 ids correspond to creation of globs, twice per glob -
for glob name, and glob stringification magic.
Here are explanations for other Ids above:
717
-
Creates bigger XPV* structures. In the case above, it creates 3 AVs
per subroutine, one for a list of lexical variable names, one for a scratchpad (which
contains lexical variables and targets), and one for the array of scratchpads
needed for recursion.
It also creates a GV and a CV per subroutine, all called from
start_subparse().
002
-
Creates a C array corresponding to the AV of scratchpads and the
scratchpad itself. The first fake entry of this scratchpad is created though the
subroutine itself is not defined yet.
It also creates C arrays to keep data for the stash. This is one HV, but it grows;
thus, there are 4 big allocations: the big chunks are not freed, but are kept as
additional arenas for SV allocations.
054
-
Creates a HEK for the name of the glob for the subroutine. This name is a
key in a stash.
Big allocations with this Id correspond to allocations of new arenas to keep HE.
602
- Creates a
GP for the glob for the subroutine.
702
- Creates the
MAGIC for the glob for the subroutine.
704
- Creates arenas which keep SVs.
If Perl is run with -DL option, then warn()s that start with `!' behave specially.
They print a list of categories of memory allocations, and statistics of allocations of
different sizes for these categories.
If warn() string starts with
!!!
- print changed categories only, print the differences in counts of allocations.
!!
- print grown categories only; print the absolute values of counts, and totals.
!
- print nonempty categories, print the absolute values of counts and totals.
If an extension or external library does not use the Perl API to allocate memory, such
allocations are not counted.
|
|
