linux内核中的内存分配睡眠问题

structpage * fastcall __alloc_pages(unsignedintgfp_mask, unsignedintorder,structzonelist *zonelist)

{

constintwait = gfp_mask & __GFP_WAIT;

unsignedlongmin;

structzone **zones, *z;

structpage *page;

structreclaim_state reclaim_state;

structtask_struct *p = current;

inti;

intalloc_type;

intdo_retry;

intcan_try_harder;//这个can_try_harder很重要，见下面初始化

might_sleep_if(wait);

can_try_harder = (unlikely(rt_task(p)) && !in_interrupt()) || !wait;

zones = zonelist->zones;/* the list of zones suitable for gfp_mask */

if(unlikely(zones[0] == NULL)) {

returnNULL;

}

alloc_type = zone_idx(zones[0]);

for(i = 0; (z = zones[i]) != NULL; i++) {

min = z->pages_low + (1< protection[alloc_type];

if(z->free_pages < min)

continue;

page = buffered_rmqueue(z, order, gfp_mask);

if(page)

gotogot_pg;

}

for(i = 0; (z = zones[i]) != NULL; i++)

wakeup_kswapd(z);//这个wakeup并不能引起进程切换，稍后解释

for(i = 0; (z = zones[i]) != NULL; i++) {//常规分配，逐渐加大强度

min = z->pages_min;

if(gfp_mask & __GFP_HIGH)

min /= 2;

if(can_try_harder)//can_try_harder影响着内存分配是否在本zone进行

min -= min / 4;

min += (1< protection[alloc_type];

if(z->free_pages < min)//满足一定条件才进入下面的buffered_rmqueue实际分配，这对于保证空闲页面在一定范围内是很重要的。

continue;

page = buffered_rmqueue(z, order, gfp_mask);

if(page)

gotogot_pg;

}

if((p->flags & (PF_MEMALLOC | PF_MEMDIE)) && !in_interrupt()) {//特权分配，没有强度限制

for(i = 0; (z = zones[i]) != NULL; i++) {

page = buffered_rmqueue(z, order, gfp_mask);

if(page)

gotogot_pg;

}

gotonopage;

}

if(!wait)//如果不能睡眠等待，比如在中断中，则直接退出此次分配

gotonopage;

rebalance://平衡内存

p->flags |= PF_MEMALLOC;

reclaim_state.reclaimed_slab = 0;

p->reclaim_state = &reclaim_state;

try_to_free_pages(zones, gfp_mask, order);//这个函数中有显式的睡眠

p->reclaim_state = NULL;

p->flags &= ~PF_MEMALLOC;

for(i = 0; (z = zones[i]) != NULL; i++) {

min = z->pages_min;

if(gfp_mask & __GFP_HIGH)

min /= 2;

if(can_try_harder)

min -= min / 4;

min += (1< protection[alloc_type];

if(z->free_pages < min)

continue;

page = buffered_rmqueue(z, order, gfp_mask);

if(page)

gotogot_pg;

}

do_retry = 0;

if(!(gfp_mask & __GFP_NORETRY)) {

if((order <= 3) || (gfp_mask & __GFP_REPEAT))

do_retry = 1;

if(gfp_mask & __GFP_NOFAIL)

do_retry = 1;

}

if(do_retry) {

blk_congestion_wait(WRITE, HZ/50);

gotorebalance;

}

nopage:

if(!(gfp_mask & __GFP_NOWARN) && printk_ratelimit()) {

printk(KERN_WARNING"%s: page allocation failure."

" order:%d, mode:0x%x\n",

p->comm, order, gfp_mask);

dump_stack();

}

returnNULL;

got_pg:

zone_statistics(zonelist, z);

kernel_map_pages(page, 1 << order, 1);

returnpage;

}

void*vmalloc(unsignedlongsize)

{

return__vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL);

}

#define GFP_KERNEL (__GFP_WAIT | __GFP_IO | __GFP_FS)