* [PATCH -next] mm/hotplug: fix an imbalance with DEBUG_PAGEALLOC
@ 2019-03-01 22:08 Qian Cai
2019-03-06 11:44 ` Michal Hocko
0 siblings, 1 reply; 2+ messages in thread
From: Qian Cai @ 2019-03-01 22:08 UTC (permalink / raw)
To: akpm; +Cc: mhocko, benh, paulus, mpe, linux-mm, linux-kernel, Qian Cai
When onlining a memory block with DEBUG_PAGEALLOC, it unmaps the pages
in the block from kernel, However, it does not map those pages while
offlining at the beginning. As the result, it triggers a panic below
while onlining on ppc64le as it checks if the pages are mapped before
unmapping. However, the imbalance exists for all arches where
double-unmappings could happen. Therefore, let kernel map those pages in
generic_online_page() before they have being freed into the page
allocator for the first time where it will set the page count to one.
On the other hand, it works fine during the boot, because at least for
IBM POWER8, it does,
early_setup
early_init_mmu
harsh__early_init_mmu
htab_initialize [1]
htab_bolt_mapping [2]
where it effectively map all memblock regions just like
kernel_map_linear_page(), so later mem_init() -> memblock_free_all()
will unmap them just fine without any imbalance. On other arches without
this imbalance checking, it still unmap them once at the most.
[1]
for_each_memblock(memory, reg) {
base = (unsigned long)__va(reg->base);
size = reg->size;
DBG("creating mapping for region: %lx..%lx (prot: %lx)\n",
base, size, prot);
BUG_ON(htab_bolt_mapping(base, base + size, __pa(base),
prot, mmu_linear_psize, mmu_kernel_ssize));
}
[2] linear_map_hash_slots[paddr >> PAGE_SHIFT] = ret | 0x80;
kernel BUG at arch/powerpc/mm/hash_utils_64.c:1815!
Oops: Exception in kernel mode, sig: 5 [#1]
LE SMP NR_CPUS=256 DEBUG_PAGEALLOC NUMA pSeries
CPU: 2 PID: 4298 Comm: bash Not tainted 5.0.0-rc7+ #15
NIP: c000000000062670 LR: c00000000006265c CTR: 0000000000000000
REGS: c0000005bf8a75b0 TRAP: 0700 Not tainted (5.0.0-rc7+)
MSR: 800000000282b033 <SF,VEC,VSX,EE,FP,ME,IR,DR,RI,LE> CR: 28422842
XER: 00000000
CFAR: c000000000804f44 IRQMASK: 1
GPR00: c00000000006265c c0000005bf8a7840 c000000001518200 c0000000013cbcc8
GPR04: 0000000000080004 0000000000000000 00000000ccc457e0 c0000005c4e341d8
GPR08: 0000000000000000 0000000000000001 c000000007f4f800 0000000000000001
GPR12: 0000000000002200 c000000007f4e100 0000000000000000 0000000139c29710
GPR16: 0000000139c29714 0000000139c29788 c0000000013cbcc8 0000000000000000
GPR20: 0000000000034000 c0000000016e05e8 0000000000000000 0000000000000001
GPR24: 0000000000bf50d9 800000000000018e 0000000000000000 c0000000016e04b8
GPR28: f000000000d00040 0000006420a2f217 f000000000d00000 00ea1b2170340000
NIP [c000000000062670] __kernel_map_pages+0x2e0/0x4f0
LR [c00000000006265c] __kernel_map_pages+0x2cc/0x4f0
Call Trace:
[c0000005bf8a7840] [c00000000006265c] __kernel_map_pages+0x2cc/0x4f0
(unreliable)
[c0000005bf8a78d0] [c00000000028c4a0] free_unref_page_prepare+0x2f0/0x4d0
[c0000005bf8a7930] [c000000000293144] free_unref_page+0x44/0x90
[c0000005bf8a7970] [c00000000037af24] __online_page_free+0x84/0x110
[c0000005bf8a79a0] [c00000000037b6e0] online_pages_range+0xc0/0x150
[c0000005bf8a7a00] [c00000000005aaa8] walk_system_ram_range+0xc8/0x120
[c0000005bf8a7a50] [c00000000037e710] online_pages+0x280/0x5a0
[c0000005bf8a7b40] [c0000000006419e4] memory_subsys_online+0x1b4/0x270
[c0000005bf8a7bb0] [c000000000616720] device_online+0xc0/0xf0
[c0000005bf8a7bf0] [c000000000642570] state_store+0xc0/0x180
[c0000005bf8a7c30] [c000000000610b2c] dev_attr_store+0x3c/0x60
[c0000005bf8a7c50] [c0000000004c0a50] sysfs_kf_write+0x70/0xb0
[c0000005bf8a7c90] [c0000000004bf40c] kernfs_fop_write+0x10c/0x250
[c0000005bf8a7ce0] [c0000000003e4b18] __vfs_write+0x48/0x240
[c0000005bf8a7d80] [c0000000003e4f68] vfs_write+0xd8/0x210
[c0000005bf8a7dd0] [c0000000003e52f0] ksys_write+0x70/0x120
[c0000005bf8a7e20] [c00000000000b000] system_call+0x5c/0x70
Instruction dump:
7fbd5278 7fbd4a78 3e42ffeb 7bbd0640 3a523ac8 7e439378 487a2881 60000000
e95505f0 7e6aa0ae 6a690080 7929c9c2 <0b090000> 7f4aa1ae 7e439378 487a28dd
Signed-off-by: Qian Cai <cai@lca.pw>
---
mm/memory_hotplug.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index c4f59ac21014..2a778602a821 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -661,6 +661,7 @@ EXPORT_SYMBOL_GPL(__online_page_free);
static void generic_online_page(struct page *page, unsigned int order)
{
+ kernel_map_pages(page, 1 << order, 1);
__free_pages_core(page, order);
totalram_pages_add(1UL << order);
#ifdef CONFIG_HIGHMEM
--
2.17.2 (Apple Git-113)
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH -next] mm/hotplug: fix an imbalance with DEBUG_PAGEALLOC
2019-03-01 22:08 [PATCH -next] mm/hotplug: fix an imbalance with DEBUG_PAGEALLOC Qian Cai
@ 2019-03-06 11:44 ` Michal Hocko
0 siblings, 0 replies; 2+ messages in thread
From: Michal Hocko @ 2019-03-06 11:44 UTC (permalink / raw)
To: Qian Cai; +Cc: akpm, benh, paulus, mpe, linux-mm, linux-kernel
On Fri 01-03-19 17:08:14, Qian Cai wrote:
> When onlining a memory block with DEBUG_PAGEALLOC, it unmaps the pages
> in the block from kernel, However, it does not map those pages while
> offlining at the beginning. As the result, it triggers a panic below
> while onlining on ppc64le as it checks if the pages are mapped before
> unmapping. However, the imbalance exists for all arches where
> double-unmappings could happen. Therefore, let kernel map those pages in
> generic_online_page() before they have being freed into the page
> allocator for the first time where it will set the page count to one.
OK, hooking into generic_online_page makes much more sense than the
previous attempt (inside offlining path).
> On the other hand, it works fine during the boot, because at least for
> IBM POWER8, it does,
>
> early_setup
> early_init_mmu
> harsh__early_init_mmu
> htab_initialize [1]
> htab_bolt_mapping [2]
>
> where it effectively map all memblock regions just like
> kernel_map_linear_page(), so later mem_init() -> memblock_free_all()
> will unmap them just fine without any imbalance. On other arches without
> this imbalance checking, it still unmap them once at the most.
>
> [1]
> for_each_memblock(memory, reg) {
> base = (unsigned long)__va(reg->base);
> size = reg->size;
>
> DBG("creating mapping for region: %lx..%lx (prot: %lx)\n",
> base, size, prot);
>
> BUG_ON(htab_bolt_mapping(base, base + size, __pa(base),
> prot, mmu_linear_psize, mmu_kernel_ssize));
> }
>
> [2] linear_map_hash_slots[paddr >> PAGE_SHIFT] = ret | 0x80;
>
> kernel BUG at arch/powerpc/mm/hash_utils_64.c:1815!
> Oops: Exception in kernel mode, sig: 5 [#1]
> LE SMP NR_CPUS=256 DEBUG_PAGEALLOC NUMA pSeries
> CPU: 2 PID: 4298 Comm: bash Not tainted 5.0.0-rc7+ #15
> NIP: c000000000062670 LR: c00000000006265c CTR: 0000000000000000
> REGS: c0000005bf8a75b0 TRAP: 0700 Not tainted (5.0.0-rc7+)
> MSR: 800000000282b033 <SF,VEC,VSX,EE,FP,ME,IR,DR,RI,LE> CR: 28422842
> XER: 00000000
> CFAR: c000000000804f44 IRQMASK: 1
> GPR00: c00000000006265c c0000005bf8a7840 c000000001518200 c0000000013cbcc8
> GPR04: 0000000000080004 0000000000000000 00000000ccc457e0 c0000005c4e341d8
> GPR08: 0000000000000000 0000000000000001 c000000007f4f800 0000000000000001
> GPR12: 0000000000002200 c000000007f4e100 0000000000000000 0000000139c29710
> GPR16: 0000000139c29714 0000000139c29788 c0000000013cbcc8 0000000000000000
> GPR20: 0000000000034000 c0000000016e05e8 0000000000000000 0000000000000001
> GPR24: 0000000000bf50d9 800000000000018e 0000000000000000 c0000000016e04b8
> GPR28: f000000000d00040 0000006420a2f217 f000000000d00000 00ea1b2170340000
> NIP [c000000000062670] __kernel_map_pages+0x2e0/0x4f0
> LR [c00000000006265c] __kernel_map_pages+0x2cc/0x4f0
> Call Trace:
> [c0000005bf8a7840] [c00000000006265c] __kernel_map_pages+0x2cc/0x4f0
> (unreliable)
> [c0000005bf8a78d0] [c00000000028c4a0] free_unref_page_prepare+0x2f0/0x4d0
> [c0000005bf8a7930] [c000000000293144] free_unref_page+0x44/0x90
> [c0000005bf8a7970] [c00000000037af24] __online_page_free+0x84/0x110
> [c0000005bf8a79a0] [c00000000037b6e0] online_pages_range+0xc0/0x150
> [c0000005bf8a7a00] [c00000000005aaa8] walk_system_ram_range+0xc8/0x120
> [c0000005bf8a7a50] [c00000000037e710] online_pages+0x280/0x5a0
> [c0000005bf8a7b40] [c0000000006419e4] memory_subsys_online+0x1b4/0x270
> [c0000005bf8a7bb0] [c000000000616720] device_online+0xc0/0xf0
> [c0000005bf8a7bf0] [c000000000642570] state_store+0xc0/0x180
> [c0000005bf8a7c30] [c000000000610b2c] dev_attr_store+0x3c/0x60
> [c0000005bf8a7c50] [c0000000004c0a50] sysfs_kf_write+0x70/0xb0
> [c0000005bf8a7c90] [c0000000004bf40c] kernfs_fop_write+0x10c/0x250
> [c0000005bf8a7ce0] [c0000000003e4b18] __vfs_write+0x48/0x240
> [c0000005bf8a7d80] [c0000000003e4f68] vfs_write+0xd8/0x210
> [c0000005bf8a7dd0] [c0000000003e52f0] ksys_write+0x70/0x120
> [c0000005bf8a7e20] [c00000000000b000] system_call+0x5c/0x70
> Instruction dump:
> 7fbd5278 7fbd4a78 3e42ffeb 7bbd0640 3a523ac8 7e439378 487a2881 60000000
> e95505f0 7e6aa0ae 6a690080 7929c9c2 <0b090000> 7f4aa1ae 7e439378 487a28dd
>
> Signed-off-by: Qian Cai <cai@lca.pw>
I can see Andrew has sent the patch to Linus already (btw. was there any
reason to rush this? It's been broken for a long time without anybody
noticing, but whatever).
Just for the reference.
Acked-by: Michal Hocko <mhocko@suse.com>
Thanks!
> ---
> mm/memory_hotplug.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index c4f59ac21014..2a778602a821 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -661,6 +661,7 @@ EXPORT_SYMBOL_GPL(__online_page_free);
>
> static void generic_online_page(struct page *page, unsigned int order)
> {
> + kernel_map_pages(page, 1 << order, 1);
> __free_pages_core(page, order);
> totalram_pages_add(1UL << order);
> #ifdef CONFIG_HIGHMEM
> --
> 2.17.2 (Apple Git-113)
--
Michal Hocko
SUSE Labs
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2019-03-06 11:44 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-03-01 22:08 [PATCH -next] mm/hotplug: fix an imbalance with DEBUG_PAGEALLOC Qian Cai
2019-03-06 11:44 ` Michal Hocko
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).