Messages in this thread | ![/](/images/icornerl.gif) | | Subject | Re: [PATCH v4] debugobjects: scale the static pool size | From | Qian Cai <> | Date | Sun, 25 Nov 2018 15:42:12 -0500 |
| |
On 11/23/18 10:01 PM, Qian Cai wrote: > > >> On Nov 22, 2018, at 4:56 PM, Thomas Gleixner <tglx@linutronix.de> wrote: >> >> On Tue, 20 Nov 2018, Qian Cai wrote: >> >> Looking deeper at that. >> >>> diff --git a/lib/debugobjects.c b/lib/debugobjects.c >>> index 70935ed91125..140571aa483c 100644 >>> --- a/lib/debugobjects.c >>> +++ b/lib/debugobjects.c >>> @@ -23,9 +23,81 @@ >>> #define ODEBUG_HASH_BITS 14 >>> #define ODEBUG_HASH_SIZE (1 << ODEBUG_HASH_BITS) >>> >>> -#define ODEBUG_POOL_SIZE 1024 >>> +#define ODEBUG_DEFAULT_POOL 512 >>> #define ODEBUG_POOL_MIN_LEVEL 256 >>> >>> +/* >>> + * Some debug objects are allocated during the early boot. Enabling some options >>> + * like timers or workqueue objects may increase the size required significantly >>> + * with large number of CPUs. For example (as today, 20 Nov. 2018), >>> + * >>> + * No. CPUs x 2 (worker pool) objects: >>> + * >>> + * start_kernel >>> + * workqueue_init_early >>> + * init_worker_pool >>> + * init_timer_key >>> + * debug_object_init >>> + * >>> + * No. CPUs objects (CONFIG_HIGH_RES_TIMERS): >>> + * >>> + * sched_init >>> + * hrtick_rq_init >>> + * hrtimer_init >>> + * >>> + * CONFIG_DEBUG_OBJECTS_WORK: >>> + * No. CPUs x 6 (workqueue) objects: >>> + * >>> + * workqueue_init_early >>> + * alloc_workqueue >>> + * __alloc_workqueue_key >>> + * alloc_and_link_pwqs >>> + * init_pwq >>> + * >>> + * Also, plus No. CPUs objects: >>> + * >>> + * perf_event_init >>> + * __init_srcu_struct >>> + * init_srcu_struct_fields >>> + * init_srcu_struct_nodes >>> + * __init_work >> >> None of the things are actually used or required _BEFORE_ >> debug_objects_mem_init() is invoked. >> >> The reason why the call is at this place in start_kernel() is >> historical. It's because back in the days when debugobjects were added the >> memory allocator was enabled way later than today. So we can just move the >> debug_objects_mem_init() call right before sched_init() I think. > > Well, now that kmemleak_init() seems complains that debug_objects_mem_init() > is called before it. > > [ 0.078805] kmemleak: Cannot insert 0xc000000dff930000 into the object search tree (overlaps existing) > [ 0.078860] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.20.0-rc3+ #3 > [ 0.078883] Call Trace: > [ 0.078904] [c000000001c8fcd0] [c000000000c96b34] dump_stack+0xe8/0x164 (unreliable) > [ 0.078935] [c000000001c8fd20] [c000000000486e84] create_object+0x344/0x380 > [ 0.078962] [c000000001c8fde0] [c000000000489544] early_alloc+0x108/0x1f8 > [ 0.078989] [c000000001c8fe20] [c00000000109738c] kmemleak_init+0x1d8/0x3d4 > [ 0.079016] [c000000001c8ff00] [c000000001054028] start_kernel+0x5c0/0x6f8 > [ 0.079043] [c000000001c8ff90] [c00000000000ae7c] start_here_common+0x1c/0x520 > [ 0.079070] kmemleak: Kernel memory leak detector disabled > [ 0.079091] kmemleak: Object 0xc000000ffd587b68 (size 40): > [ 0.079112] kmemleak: comm "swapper/0", pid 0, jiffies 4294937299 > [ 0.079135] kmemleak: min_count = -1 > [ 0.079153] kmemleak: count = 0 > [ 0.079170] kmemleak: flags = 0x5 > [ 0.079188] kmemleak: checksum = 0 > [ 0.079206] kmemleak: backtrace: > [ 0.079227] __debug_object_init+0x688/0x700 > [ 0.079250] debug_object_activate+0x1e0/0x350 > [ 0.079272] __call_rcu+0x60/0x430 > [ 0.079292] put_object+0x60/0x80 > [ 0.079311] kmemleak_init+0x2cc/0x3d4 > [ 0.079331] start_kernel+0x5c0/0x6f8 > [ 0.079351] start_here_common+0x1c/0x520 > [ 0.079380] kmemleak: Early log backtrace: > [ 0.079399] memblock_alloc_try_nid_raw+0x90/0xcc > [ 0.079421] sparse_init_nid+0x144/0x51c > [ 0.079440] sparse_init+0x1a0/0x238 > [ 0.079459] initmem_init+0x1d8/0x25c > [ 0.079498] setup_arch+0x3e0/0x464 > [ 0.079517] start_kernel+0xa4/0x6f8 > [ 0.079536] start_here_common+0x1c/0x520 >
So this is an chicken-egg problem. Debug objects need kmemleak_init() first, so it can make use of kmemleak_ignore() for all debug objects in order to avoid the overlapping like the above.
while (obj_pool_free < debug_objects_pool_min_level) {
new = kmem_cache_zalloc(obj_cache, gfp); if (!new) return;
kmemleak_ignore(new);
However, there seems no way to move kmemleak_init() together this early in start_kernel() just before vmalloc_init() [1] because it looks like it depends on things like workqueue (schedule_work(&cleanup_work)) and rcu. Hence, it needs to be after workqueue_init_early() and rcu_init()
Given that, maybe the best outcome is to stick to the alternative approach that works [1] rather messing up with the order of debug_objects_mem_init() in start_kernel() which seems tricky. What do you think?
[1] https://goo.gl/18N78g [2] https://goo.gl/My6ig6
| ![\](/images/icornerr.gif) |