From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michal Hocko Date: Tue, 15 Sep 2020 12:19:05 +0000 Subject: Re: [PATCH v3 2/3] mm: don't rely on system state to detect hot-plug operations Message-Id: <20200915121905.GE4649@dhcp22.suse.cz> List-Id: References: <20200915094143.79181-1-ldufour@linux.ibm.com> <20200915094143.79181-3-ldufour@linux.ibm.com> In-Reply-To: <20200915094143.79181-3-ldufour@linux.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: quoted-printable To: Laurent Dufour Cc: akpm@linux-foundation.org, David Hildenbrand , Oscar Salvador , Greg Kroah-Hartman , linux-mm@kvack.org, "Rafael J . Wysocki" , nathanl@linux.ibm.com, cheloha@linux.ibm.com, Tony Luck , Fenghua Yu , linux-ia64@vger.kernel.org, linux-kernel@vger.kernel.org, stable@vger.kernel.org On Tue 15-09-20 11:41:42, Laurent Dufour wrote: > In register_mem_sect_under_node() the system_state=E2=80=99s value is che= cked to > detect whether the call is made during boot time or during an hot-plug > operation. Unfortunately, that check against SYSTEM_BOOTING is wrong > because regular memory is registered at SYSTEM_SCHEDULING state. In > addition, memory hot-plug operation can be triggered at this system state > by the ACPI [1]. So checking against the system state is not enough. >=20 > The consequence is that on system with interleaved node's ranges like thi= s: > Early memory node ranges > node 1: [mem 0x0000000000000000-0x000000011fffffff] > node 2: [mem 0x0000000120000000-0x000000014fffffff] > node 1: [mem 0x0000000150000000-0x00000001ffffffff] > node 0: [mem 0x0000000200000000-0x000000048fffffff] > node 2: [mem 0x0000000490000000-0x00000007ffffffff] >=20 > This can be seen on PowerPC LPAR after multiple memory hot-plug and > hot-unplug operations are done. At the next reboot the node's memory rang= es > can be interleaved and since the call to link_mem_sections() is made in > topology_init() while the system is in the SYSTEM_SCHEDULING state, the > node's id is not checked, and the sections registered to multiple nodes: >=20 > $ ls -l /sys/devices/system/memory/memory21/node* > total 0 > lrwxrwxrwx 1 root root 0 Aug 24 05:27 node1 -> ../../node/node1 > lrwxrwxrwx 1 root root 0 Aug 24 05:27 node2 -> ../../node/node2 >=20 > In that case, the system is able to boot but if later one of theses memory > blocks is hot-unplugged and then hot-plugged, the sysfs inconsistency is > detected and this is triggering a BUG_ON(): >=20 > ------------[ cut here ]------------ > kernel BUG at /Users/laurent/src/linux-ppc/mm/memory_hotplug.c:1084! > Oops: Exception in kernel mode, sig: 5 [#1] > LE PAGE_SIZEdK MMU=3DHash SMP NR_CPUS 48 NUMA pSeries > Modules linked in: rpadlpar_io rpaphp pseries_rng rng_core vmx_crypto gf1= 28mul binfmt_misc ip_tables x_tables xfs libcrc32c crc32c_vpmsum autofs4 > CPU: 8 PID: 10256 Comm: drmgr Not tainted 5.9.0-rc1+ #25 > NIP: c000000000403f34 LR: c000000000403f2c CTR: 0000000000000000 > REGS: c0000004876e3660 TRAP: 0700 Not tainted (5.9.0-rc1+) > MSR: 800000000282b033 CR: 24000448 X= ER: 20040000 > CFAR: c000000000846d20 IRQMASK: 0 > GPR00: c000000000403f2c c0000004876e38f0 c0000000012f6f00 ffffffffffffffef > GPR04: 0000000000000227 c0000004805ae680 0000000000000000 00000004886f0000 > GPR08: 0000000000000226 0000000000000003 0000000000000002 fffffffffffffffd > GPR12: 0000000088000484 c00000001ec96280 0000000000000000 0000000000000000 > GPR16: 0000000000000000 0000000000000000 0000000000000004 0000000000000003 > GPR20: c00000047814ffe0 c0000007ffff7c08 0000000000000010 c0000000013332c8 > GPR24: 0000000000000000 c0000000011f6cc0 0000000000000000 0000000000000000 > GPR28: ffffffffffffffef 0000000000000001 0000000150000000 0000000010000000 > NIP [c000000000403f34] add_memory_resource+0x244/0x340 > LR [c000000000403f2c] add_memory_resource+0x23c/0x340 > Call Trace: > [c0000004876e38f0] [c000000000403f2c] add_memory_resource+0x23c/0x340 (un= reliable) > [c0000004876e39c0] [c00000000040408c] __add_memory+0x5c/0xf0 > [c0000004876e39f0] [c0000000000e2b94] dlpar_add_lmb+0x1b4/0x500 > [c0000004876e3ad0] [c0000000000e3888] dlpar_memory+0x1f8/0xb80 > [c0000004876e3b60] [c0000000000dc0d0] handle_dlpar_errorlog+0xc0/0x190 > [c0000004876e3bd0] [c0000000000dc398] dlpar_store+0x198/0x4a0 > [c0000004876e3c90] [c00000000072e630] kobj_attr_store+0x30/0x50 > [c0000004876e3cb0] [c00000000051f954] sysfs_kf_write+0x64/0x90 > [c0000004876e3cd0] [c00000000051ee40] kernfs_fop_write+0x1b0/0x290 > [c0000004876e3d20] [c000000000438dd8] vfs_write+0xe8/0x290 > [c0000004876e3d70] [c0000000004391ac] ksys_write+0xdc/0x130 > [c0000004876e3dc0] [c000000000034e40] system_call_exception+0x160/0x270 > [c0000004876e3e20] [c00000000000d740] system_call_common+0xf0/0x27c > Instruction dump: > 48442e35 60000000 0b030000 3cbe0001 7fa3eb78 7bc48402 38a5fffe 7ca5fa14 > 78a58402 48442db1 60000000 7c7c1b78 <0b030000> 7f23cb78 4bda371d 60000000 > ---[ end trace 562fd6c109cd0fb2 ]--- >=20 > This patch addresses the root cause by not relying on the system_state > value to detect whether the call is due to a hot-plug operation. An extra > parameter is added to link_mem_sections() detailing whether the operation > is due to a hot-plug operation. >=20 > [1] According to Oscar Salvador, using this qemu command line, ACPI memory > hotplug operations are raised at SYSTEM_SCHEDULING state: >=20 > $QEMU -enable-kvm -machine pc -smp 4,sockets=3D4,cores=3D1,threads=3D1 -c= pu host -monitor pty \ > -m size=3D$MEM,slots%5,maxmemB94967296k \ > -numa node,nodeid=3D0,cpus=3D0-3,memQ2 -numa node,nodeid=3D1,memQ= 2 \ > -object memory-backend-ram,id=3Dmemdimm0,size=134217728 -device p= c-dimm,node=3D0,memdev=3Dmemdimm0,id=3Ddimm0,slot=3D0 \ > -object memory-backend-ram,id=3Dmemdimm1,size=134217728 -device p= c-dimm,node=3D0,memdev=3Dmemdimm1,id=3Ddimm1,slot=3D1 \ > -object memory-backend-ram,id=3Dmemdimm2,size=134217728 -device p= c-dimm,node=3D0,memdev=3Dmemdimm2,id=3Ddimm2,slot=3D2 \ > -object memory-backend-ram,id=3Dmemdimm3,size=134217728 -device p= c-dimm,node=3D0,memdev=3Dmemdimm3,id=3Ddimm3,slot=3D3 \ > -object memory-backend-ram,id=3Dmemdimm4,size=134217728 -device p= c-dimm,node=3D1,memdev=3Dmemdimm4,id=3Ddimm4,slot=3D4 \ > -object memory-backend-ram,id=3Dmemdimm5,size=134217728 -device p= c-dimm,node=3D1,memdev=3Dmemdimm5,id=3Ddimm5,slot=3D5 \ > -object memory-backend-ram,id=3Dmemdimm6,size=134217728 -device p= c-dimm,node=3D1,memdev=3Dmemdimm6,id=3Ddimm6,slot=3D6 \ >=20 > Fixes: 4fbce633910e ("mm/memory_hotplug.c: make register_mem_sect_under_n= ode() a callback of walk_memory_range()") > Signed-off-by: Laurent Dufour > Reviewed-by: David Hildenbrand > Cc: stable@vger.kernel.org > Cc: Greg Kroah-Hartman > Cc: "Rafael J. Wysocki" > Cc: Andrew Morton > Cc: Michal Hocko > Cc: Oscar Salvador Acked-by: Michal Hocko > --- > drivers/base/node.c | 85 ++++++++++++++++++++++++++++---------------- > include/linux/node.h | 11 +++--- > mm/memory_hotplug.c | 3 +- > 3 files changed, 64 insertions(+), 35 deletions(-) >=20 > diff --git a/drivers/base/node.c b/drivers/base/node.c > index 508b80f6329b..50af16e68d98 100644 > --- a/drivers/base/node.c > +++ b/drivers/base/node.c > @@ -761,14 +761,36 @@ static int __ref get_nid_for_pfn(unsigned long pfn) > return pfn_to_nid(pfn); > } > =20 > +static int do_register_memory_block_under_node(int nid, > + struct memory_block *mem_blk) > +{ > + int ret; > + > + /* > + * If this memory block spans multiple nodes, we only indicate > + * the last processed node. > + */ > + mem_blk->nid =3D nid; > + > + ret =3D sysfs_create_link_nowarn(&node_devices[nid]->dev.kobj, > + &mem_blk->dev.kobj, > + kobject_name(&mem_blk->dev.kobj)); > + if (ret) > + return ret; > + > + return sysfs_create_link_nowarn(&mem_blk->dev.kobj, > + &node_devices[nid]->dev.kobj, > + kobject_name(&node_devices[nid]->dev.kobj)); > +} > + > /* register memory section under specified node if it spans that node */ > -static int register_mem_sect_under_node(struct memory_block *mem_blk, > - void *arg) > +static int register_mem_block_under_node_early(struct memory_block *mem_= blk, > + void *arg) > { > unsigned long memory_block_pfns =3D memory_block_size_bytes() / PAGE_SI= ZE; > unsigned long start_pfn =3D section_nr_to_pfn(mem_blk->start_section_nr= ); > unsigned long end_pfn =3D start_pfn + memory_block_pfns - 1; > - int ret, nid =3D *(int *)arg; > + int nid =3D *(int *)arg; > unsigned long pfn; > =20 > for (pfn =3D start_pfn; pfn <=3D end_pfn; pfn++) { > @@ -785,38 +807,33 @@ static int register_mem_sect_under_node(struct memo= ry_block *mem_blk, > } > =20 > /* > - * We need to check if page belongs to nid only for the boot > - * case, during hotplug we know that all pages in the memory > - * block belong to the same node. > - */ > - if (system_state =3D SYSTEM_BOOTING) { > - page_nid =3D get_nid_for_pfn(pfn); > - if (page_nid < 0) > - continue; > - if (page_nid !=3D nid) > - continue; > - } > - > - /* > - * If this memory block spans multiple nodes, we only indicate > - * the last processed node. > + * We need to check if page belongs to nid only at the boot > + * case because node's ranges can be interleaved. > */ > - mem_blk->nid =3D nid; > - > - ret =3D sysfs_create_link_nowarn(&node_devices[nid]->dev.kobj, > - &mem_blk->dev.kobj, > - kobject_name(&mem_blk->dev.kobj)); > - if (ret) > - return ret; > + page_nid =3D get_nid_for_pfn(pfn); > + if (page_nid < 0) > + continue; > + if (page_nid !=3D nid) > + continue; > =20 > - return sysfs_create_link_nowarn(&mem_blk->dev.kobj, > - &node_devices[nid]->dev.kobj, > - kobject_name(&node_devices[nid]->dev.kobj)); > + return do_register_memory_block_under_node(nid, mem_blk); > } > /* mem section does not span the specified node */ > return 0; > } > =20 > +/* > + * During hotplug we know that all pages in the memory block belong to t= he same > + * node. > + */ > +static int register_mem_block_under_node_hotplug(struct memory_block *me= m_blk, > + void *arg) > +{ > + int nid =3D *(int *)arg; > + > + return do_register_memory_block_under_node(nid, mem_blk); > +} > + > /* > * Unregister a memory block device under the node it spans. Memory bloc= ks > * with multiple nodes cannot be offlined and therefore also never be re= moved. > @@ -832,11 +849,19 @@ void unregister_memory_block_under_nodes(struct mem= ory_block *mem_blk) > kobject_name(&node_devices[mem_blk->nid]->dev.kobj)); > } > =20 > -int link_mem_sections(int nid, unsigned long start_pfn, unsigned long en= d_pfn) > +int link_mem_sections(int nid, unsigned long start_pfn, unsigned long en= d_pfn, > + enum meminit_context context) > { > + walk_memory_blocks_func_t func; > + > + if (context =3D MEMINIT_HOTPLUG) > + func =3D register_mem_block_under_node_hotplug; > + else > + func =3D register_mem_block_under_node_early; > + > return walk_memory_blocks(PFN_PHYS(start_pfn), > PFN_PHYS(end_pfn - start_pfn), (void *)&nid, > - register_mem_sect_under_node); > + func); > } > =20 > #ifdef CONFIG_HUGETLBFS > diff --git a/include/linux/node.h b/include/linux/node.h > index 4866f32a02d8..014ba3ab2efd 100644 > --- a/include/linux/node.h > +++ b/include/linux/node.h > @@ -99,11 +99,13 @@ extern struct node *node_devices[]; > typedef void (*node_registration_func_t)(struct node *); > =20 > #if defined(CONFIG_MEMORY_HOTPLUG_SPARSE) && defined(CONFIG_NUMA) > -extern int link_mem_sections(int nid, unsigned long start_pfn, > - unsigned long end_pfn); > +int link_mem_sections(int nid, unsigned long start_pfn, > + unsigned long end_pfn, > + enum meminit_context context); > #else > static inline int link_mem_sections(int nid, unsigned long start_pfn, > - unsigned long end_pfn) > + unsigned long end_pfn, > + enum meminit_context context) > { > return 0; > } > @@ -128,7 +130,8 @@ static inline int register_one_node(int nid) > if (error) > return error; > /* link memory sections under this node */ > - error =3D link_mem_sections(nid, start_pfn, end_pfn); > + error =3D link_mem_sections(nid, start_pfn, end_pfn, > + MEMINIT_EARLY); > } > =20 > return error; > diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c > index fc25886ad719..03df20078827 100644 > --- a/mm/memory_hotplug.c > +++ b/mm/memory_hotplug.c > @@ -1080,7 +1080,8 @@ int __ref add_memory_resource(int nid, struct resou= rce *res) > } > =20 > /* link memory sections under this node.*/ > - ret =3D link_mem_sections(nid, PFN_DOWN(start), PFN_UP(start + size - 1= )); > + ret =3D link_mem_sections(nid, PFN_DOWN(start), PFN_UP(start + size - 1= ), > + MEMINIT_HOTPLUG); > BUG_ON(ret); > =20 > /* create new memmap entry */ > --=20 > 2.28.0 >=20 --=20 Michal Hocko SUSE Labs