Memory not being released

[thecrypticace]

It appears that allocated space is not being freed resulting in large memory usage. I believe entire object graphs are staying around when stored in sets.

I've tested these classes that all seem to have the problem:

• Set
• Map
• Stack
• Queue
• Deque

Notes: Neither calling clear() on the set nor the use of gc_collect_cycles() frees any memory.

Example reproduction:

$original = memory_get_usage(true);

for ($i = 0; $i < 10000000; $i++) {
    new \Ds\Set(["A", "B"]);
}

echo memory_get_usage(true) - $original, " additional bytes allocated\n";

Output:

79691776 additional bytes allocated

Environment:

• PHP DS compiled from master
• PHP 7.0.9 64-bit
• macOS 10.11.6 / Ubuntu 14.04

Edit: 10x iteration count + output changes to showcase larger memory footprint

[rtheunissen]

It's useful to run this using a debug build, which shows that there is a leak, and that the allocation was made here: https://github.com/php-ds/extension/blob/master/src/ds/ds_set.c#L12

ds_set_t *set = ecalloc(1, sizeof(ds_set_t));

We're obviously not freeing sets properly, and we should also check for memory leaks in CI. I'll open an issue to discuss to best way to do that.

[rtheunissen]

Fixed in e937fb5

[thecrypticace]

It's useful to run this using a debug build,

That's good to know. I might try setting one up on my end to test things.

Fixed in …

It appears that as soon as a union, intersection, xor, etc… are performed memory will still be leaked.

While C is not my strong suit I know my way around some of it. I might try my hand at and fixing some of these issues. I'm sure PRs fixing (or — attempting to) could be helpful?

[rtheunissen]

While C is not my strong suit I know my way around some of it. I might try my hand at and fixing some of these issues. I'm sure PRs fixing (or — attempting to) could be helpful?

Your detective work has been incredibly helpful to the point where I get excited when I see you've created a new issue or PR.

PR's fixing would definitely be helpful though, would be cool to have someone else dig into the C as well. It's very satisfying to fix these.

It appears that as soon as a union, intersection, xor, etc… are performed memory will still be leaked.

How can you tell? 🤔

[thecrypticace]

I boiled it down to a test case:

$original = memory_get_usage(true);

for ($i = 0; $i < 10e4; $i++) {
    $a = new \Ds\Set([]);
    $b = new \Ds\Set([]);
    $a->xor($b);
}

echo memory_get_usage(true) - $original, " additional bytes allocated\n";

Output:

60817408 additional bytes allocated

Aside: I've got another issue where php-ds freezes related to sets containing Hashable objects containing sets containing Hashable objects (… and then turtles?). I'll open another issue with the details.

[rtheunissen]

Debug build output for a single loop shows:

[Thu Aug  4 17:05:39 2016]  Script:  '/path/to/leak-test.php'
/path/to/ds/src/ds/ds_set.c(12) :  Freeing 0x105202078 (8 bytes), script=/path/to/leak-test.php
/path/to/php/Zend/zend_alloc.c(2510) : Actual location (location was relayed)
[Thu Aug  4 17:05:39 2016]  Script:  '/path/to/leak-test.php'
/path/to/ds/src/ds/ds_htable.c(142) :  Freeing 0x105203CC0 (32 bytes), script=/path/to/leak-test.php
/path/to/php/Zend/zend_alloc.c(2510) : Actual location (location was relayed)
[Thu Aug  4 17:05:39 2016]  Script:  '/path/to/leak-test.php'
/path/to/ds/src/ds/ds_htable.c(37) :  Freeing 0x10525ABA0 (64 bytes), script=/path/to/leak-test.php
[Thu Aug  4 17:05:39 2016]  Script:  '/path/to/leak-test.php'
/path/to/ds/src/ds/ds_htable.c(27) :  Freeing 0x105268500 (512 bytes), script=/path/to/leak-test.php
/path/to/php/Zend/zend_alloc.c(2510) : Actual location (location was relayed)
=== Total 4 memory leaks detected ===

Which collapses down to:

ds_set.c(12)
ds_htable.c(142)
ds_htable.c(37)
ds_htable.c(27)