Skip to contentSkip to conversation

kmalloc without GFP_xxx?

DISCUSSION Manfred Spraul 1 participant 1 message Jun 29, 2005

lore
Share & tools
lore
#0 lore raw
Quoted 5 lines · Denis

It struck me that kernel actually can figure out whether it's okay to sleep or not by looking at combination of (flags & __GFP_WAIT) and ((in_atomic() || irqs_disabled()) as it already does this for might_sleep() barfing:

Wrong:
- the kernel cannot figure out if a thread owns a normal spinlock(): in_atomic detects spin_lock_bh(), irqs_disabled spin_lock_irq(). But spin_lock has no global state.
- dito for get_cpu()/put_cpu users.
- dito for rcu users, or anyone else that uses preempt_disable() for whatever purpose

--
Manfred