141 Commits

Author	SHA1	Message	Date
Caleb James DeLisle	c1d3824ef1	Removed unnecessary includes and a few unused/useless files	2024-09-28 14:43:16 +00:00
Caleb James DeLisle	fd895ea982	Formally make onFree synchronous and freeing of allocators happens under lock	2024-04-25 18:59:06 +02:00
Caleb James DeLisle	e6160b456b	Make Allocator_free() always synchronous. When async freeing is needed, chain allocators.	2023-02-26 23:32:54 +01:00
Caleb James DeLisle	0af9d5a4b8	1. Adopting a root alloc is a no-op ... 2. Handle realloc(0), then realloc(non-0) 3. Logging	2022-10-17 05:16:40 +02:00
Caleb James DeLisle	afa43216ef	Rewrite of allocator.rs, also don't allocate memory if we're freeing	2022-04-14 13:48:47 +02:00
Caleb James DeLisle	d26e80a318	Move to Rust Allocator	2022-04-07 02:47:27 +02:00
Caleb James DeLisle	e2cfe83c94	1. We cannot re-arm a dead timeout because it keeps blocking the event loop ref count ... 2. Remaining memory issue with allocator teardown, ignore it for now.	2022-02-17 05:26:13 +01:00
Rogério Sampaio de Almeida	70c2a8f001	typo	2022-01-18 18:31:12 -03:00
Caleb James DeLisle	a3b7f328d9	Standardized the Rust Iface, Rust Allocator interop and Rust/C FFI	2020-11-29 12:58:38 +01:00
Caleb James DeLisle	b9e41a2521	Make sure to cast uintptr to long before passing it to a printf	2020-09-24 15:07:55 +02:00
Casper Ti. Vector	2298997092	Manual merge 79a60b9	2020-09-18 22:08:23 +02:00
Caleb James DeLisle	b6a1d3bf2f	Switch from using long to using uintptr_t because on windows, long is not pointer-size	2020-07-11 01:19:26 +02:00
Caleb James DeLisle	ca4be7578d	Rid the project of setjmp()	2020-04-28 19:58:28 +02:00
Caleb James DeLisle	9ce4bf1062	Fix cjdns tests with new version of libuv, also fix a bug which was exposed in the allocation system wherein allocating memory in an onFreeJob does the wrong thing	2019-11-25 00:20:00 +01:00
Caleb James DeLisle	40d53a658b	Deal with machines where __BIGGEST_ALIGMENT__ is bigger than sizeof(struct Allocator_Allocation_pvt)	2018-10-04 22:14:42 +02:00
Caleb James DeLisle	9b963631b5	Wrong type in printf	2018-08-04 17:38:19 +02:00
Caleb James DeLisle	db1f135a28	Shuffle around a couple fields in order to hopefully get an aligned structure	2018-08-04 17:36:40 +02:00
Caleb James DeLisle	bdcc95a833	A different approach to aligning memory allocations	2018-08-04 16:39:46 +02:00
Caleb James DeLisle	c5e41c66a7	Oops, missing one pad	2018-08-04 13:41:23 +02:00
Caleb James DeLisle	436d59060d	Align new allocations on __BIGGEST_ALIGNMENT__ instead of __SIZEOF_POINTER__ which is not reliable	2018-08-04 13:33:32 +02:00
Mada	f98c7b8136	Enable -Wmissing-prototypes compile option	2018-04-12 10:38:01 +08:00
Caleb James DeLisle	feabb1970f	Ok, never mind, this was a silly thing, it was correct in the beginning	2017-07-19 10:08:54 +02:00
Caleb James DeLisle	41efd0c895	Hopefully fixed the allocator failure to account memory properly	2017-07-18 21:55:03 +02:00
Caleb James DeLisle	7f55d9bfb7	Apply madafoo's change as-is from long ago which I managed to somehow misunderstood and refused for a really long time :(	2017-07-18 09:16:34 +02:00
Caleb James DeLisle	1881da934d	Automatic refactoring to get rid of dangerous usage of String_CONST()	2017-03-30 20:27:16 +02:00
Caleb James DeLisle	c8b4cf9a40	remove some deprecations and replace Dict_*(String_CONST()) with Dict_*C()	2017-03-30 20:15:03 +02:00
Caleb James DeLisle	dca6cdcccd	Pulling this commit and making a branch allocator-fix because this turns out to be unravelling a much deeper issue which cannot be fixed so quickly. ... Revert "Correct probably bad memory handling in allocator" This reverts commit 043b6eadd7.	2017-02-19 22:50:26 +01:00
Caleb James DeLisle	043b6eadd7	Correct probably bad memory handling in allocator	2017-02-18 01:34:40 +01:00
soredake	c49cc24869	links update	2017-01-14 00:57:12 +02:00
Caleb James DeLisle	54af8cca0c	Change Linker_require() to allow a trailing semicolon	2016-01-06 19:44:49 +01:00
Caleb James DeLisle	660ac4004b	Merge branch 'disconnect-adopted-dry-refactor' of git://github.com/gubatron/cjdns into crashey	2015-11-30 20:31:27 +01:00
gubatron	0fd96298b0	rename misleading findAllocator to disconnectAllocator	2015-11-28 18:59:39 -07:00
gubatron	35c46a6708	DRY refactor on Allocator::disconnectAdopted	2015-11-19 15:29:29 -05:00
Kubuxu	a38e58b1c7	Remove usage of Bits_memcpyConst and Bits_memmoveConst	2015-11-06 17:42:56 +01:00
Caleb James DeLisle	0000500588	Give the compiler a little bit more instruction-scheduling leeway in a function which is sometimes hot	2015-10-25 23:06:54 +01:00
Caleb James DeLisle	c5bd11ce66	small improvements to the allocator's error catching infra	2015-10-20 09:49:09 +02:00
Caleb James DeLisle	24401e7b9a	complain very loudly if the memory allocation logic is leaking in some way	2015-10-18 17:37:29 +02:00
Caleb James DeLisle	ab3a0f4452	wip	2015-03-20 13:31:46 +01:00
Caleb James DeLisle	f19fe177bb	wip	2015-03-15 22:14:58 +01:00
Caleb James DeLisle	a6afb74e25	wip	2015-03-13 09:06:28 +01:00
Caleb James DeLisle	44d393d9e7	Merge branch 'crashey' into dead-angel	2015-03-08 14:22:47 +01:00
Caleb James DeLisle	07be184ee3	try allowing clang to see Gcc_NORETURN	2015-03-08 00:53:09 +01:00
Caleb James DeLisle	2dddd5d6e5	some compilers are a bit slow -- but smart enough not to allow *that*	2015-03-08 00:46:04 +01:00
Caleb James DeLisle	27c7f234e3	some compilers are a bit slow	2015-03-08 00:43:24 +01:00
Caleb James DeLisle	315d643bfe	Support communist non-free compilers	2015-03-07 23:58:17 +01:00
Caleb James DeLisle	b8e1e3e0b9	wip	2015-03-05 09:23:40 +01:00
Caleb James DeLisle	a5117df6e3	Change Pathfinder to notify the core whenever a better path is discovered by the NodeStore, also some refactoring in the Allocator	2015-03-03 07:58:16 +01:00
Caleb James DeLisle	f8c4656984	checkpoint before running some sed	2015-02-15 22:35:31 +01:00
Caleb James DeLisle	c7eed6b146	more robust logging, although only about 100 messages can be logged at a time without drops...	2014-10-23 00:20:33 +02:00
Caleb James DeLisle	7667736795	NodeStore_brokenLink() instead of NodeStore_brokenPath()	2014-09-14 23:35:25 +02:00