This patch introduces PF_MEMALLOC_NOIO on process flag('flags' field of
'struct task_struct'), so that the flag can be set by one task
to avoid doing I/O inside memory allocation in the task's context.
The patch trys to solve one deadlock problem caused by block device,
and the problem may happen at least in the below situations:
- during block device runtime resume, if memory allocation with
GFP_KERNEL is called inside runtime resume callback of any one
of its ancestors(or the block device itself), the deadlock may be
triggered inside the memory allocation since it might not complete
until the block device becomes active and the involed page I/O finishes.
The situation is pointed out first by Alan Stern. It is not a good
approach to convert all GFP_KERNEL[1] in the path into GFP_NOIO because
several subsystems may be involved(for example, PCI, USB and SCSI may
be involved for usb mass stoarage device, network devices involved too
in the iSCSI case)
- during error handling of usb mass storage deivce, USB bus reset
will be put on the device, so there shouldn't have any
memory allocation with GFP_KERNEL during USB bus reset, otherwise
the deadlock similar with above may be triggered. Unfortunately, any
usb device may include one mass storage interface in theory, so it
requires all usb interface drivers to handle the situation. In fact,
most usb drivers don't know how to handle bus reset on the device
and don't provide .pre_set() and .post_reset() callback at all, so
USB core has to unbind and bind driver for these devices. So it
is still not practical to resort to GFP_NOIO for solving the problem.
Also the introduced solution can be used by block subsystem or block
drivers too, for example, set the PF_MEMALLOC_NOIO flag before doing
actual I/O transfer.
It is not a good idea to convert all these GFP_KERNEL in the
affected path into GFP_NOIO because these functions doing that may be
implemented as library and will be called in many other contexts.
In fact, memalloc_noio() can convert some of current static GFP_NOIO
allocation into GFP_KERNEL back in other non-affected contexts, at least
almost all GFP_NOIO in USB subsystem can be converted into GFP_KERNEL
after applying the approach and make allocation with GFP_IO
only happen in runtime resume/bus reset/block I/O transfer contexts
generally.
[1], several GFP_KERNEL allocation examples in runtime resume path
- pci subsystem
acpi_os_allocate
<-acpi_ut_allocate
<-ACPI_ALLOCATE_ZEROED
<-acpi_evaluate_object
<-__acpi_bus_set_power
<-acpi_bus_set_power
<-acpi_pci_set_power_state
<-platform_pci_set_power_state
<-pci_platform_power_transition
<-__pci_complete_power_transition
<-pci_set_power_state
<-pci_restore_standard_config
<-pci_pm_runtime_resume
- usb subsystem
usb_get_status
<-finish_port_resume
<-usb_port_resume
<-generic_resume
<-usb_resume_device
<-usb_resume_both
<-usb_runtime_resume
- some individual usb drivers
usblp, uvc, gspca, most of dvb-usb-v2 media drivers, cpia2, az6007, ....
That is just what I have found. Unfortunately, this allocation can
only be found by human being now, and there should be many not found
since any function in the resume path(call tree) may allocate memory
with GFP_KERNEL.
Cc: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>
Cc: Oliver Neukum <oneukum@xxxxxxx>
Cc: Jiri Kosina <jiri.kosina@xxxxxxxx>
Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Cc: Mel Gorman <mel@xxxxxxxxx>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: "Rafael J. Wysocki" <rjw@xxxxxxx>
Signed-off-by: Minchan Kim <minchan@xxxxxxxxxx>
Signed-off-by: Ming Lei <ming.lei@xxxxxxxxxxxxx>
---
v2:
- remove changes on 'may_writepage' and 'may_swap' because that
isn't related with the patchset, and can't introduce I/O in
allocation path if GFP_IOFS is unset, so handing 'may_swap'
and may_writepage on GFP_NOIO or GFP_NOFS should be a
mm internal thing, and let mm guys deal with that, :-).
Looks clearing the two may_XXX flag only excludes dirty pages and
anon pages for relaiming, and the behaviour should be decided
by GFP FLAG, IMO.
- unset GFP_IOFS in try_to_free_pages() path since alloc_page_buffers()
and dma_alloc_from_contiguous may drop into the path, as
pointed by KAMEZAWA Hiroyuki
v1:
- take Minchan's change to avoid the check in alloc_page hot path
- change the helpers' style into save/restore as suggested by Alan
Stern
---
include/linux/sched.h | 10 ++++++++++
mm/page_alloc.c | 10 +++++++++-
mm/vmscan.c | 12 ++++++++++++
3 files changed, 31 insertions(+), 1 deletion(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index f7a76fa..ac5234a 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1793,6 +1793,7 @@ extern void thread_group_times(struct task_struct *p, cputime_t *ut, cputime_t *
#define PF_FROZEN 0x00010000 /* frozen for system suspend */
#define PF_FSTRANS 0x00020000 /* inside a filesystem transaction */
#define PF_KSWAPD 0x00040000 /* I am kswapd */
+#define PF_MEMALLOC_NOIO 0x00080000 /* Allocating memory without IO involved */
#define PF_LESS_THROTTLE 0x00100000 /* Throttle me less: I clean memory */
#define PF_KTHREAD 0x00200000 /* I am a kernel thread */
#define PF_RANDOMIZE 0x00400000 /* randomize virtual address space */
@@ -1830,6 +1831,15 @@ extern void thread_group_times(struct task_struct *p, cputime_t *ut, cputime_t *
#define tsk_used_math(p) ((p)->flags & PF_USED_MATH)
#define used_math() tsk_used_math(current)
+#define memalloc_noio() (current->flags & PF_MEMALLOC_NOIO)
+#define memalloc_noio_save(flag) do { \
+ (flag) = current->flags & PF_MEMALLOC_NOIO; \
+ current->flags |= PF_MEMALLOC_NOIO; \
+} while (0)
+#define memalloc_noio_restore(flag) do { \
+ current->flags = (current->flags & ~PF_MEMALLOC_NOIO) | flag; \
+} while (0)
+
/*
* task->jobctl flags
*/
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 0c871fc..a7b76ae 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2630,10 +2630,18 @@ retry_cpuset:
page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, nodemask, order,
zonelist, high_zoneidx, alloc_flags,
preferred_zone, migratetype);
- if (unlikely(!page))
+ if (unlikely(!page)) {
+ /*
+ * Resume, block IO and its error handling path
+ * can deadlock because I/O on the device might not
+ * complete.
+ */
+ if (unlikely(memalloc_noio()))
+ gfp_mask &= ~GFP_IOFS;
page = __alloc_pages_slowpath(gfp_mask, order,
zonelist, high_zoneidx, nodemask,
preferred_zone, migratetype);
+ }
trace_mm_page_alloc(page, order, gfp_mask, migratetype);
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 2624edc..5bf8290 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2298,6 +2298,12 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
.gfp_mask = sc.gfp_mask,
};
+ if (unlikely(memalloc_noio())) {
+ gfp_mask &= ~GFP_IOFS;
+ sc.gfp_mask = gfp_mask;
+ shrink.gfp_mask = sc.gfp_mask;
+ }
+
throttle_direct_reclaim(gfp_mask, zonelist, nodemask);
/*
@@ -3298,6 +3304,12 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
};
unsigned long nr_slab_pages0, nr_slab_pages1;
+ if (unlikely(memalloc_noio())) {
+ gfp_mask &= ~GFP_IOFS;
+ sc.gfp_mask = gfp_mask;
+ shrink.gfp_mask = sc.gfp_mask;
+ }
+
cond_resched();
/*
* We need to be able to allocate from the reserves for RECLAIM_SWAP
--
1.7.9.5
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>