From: Mike Rapoport rppt@linux.ibm.com
Jiri Olsa reported a fault when running:
# cat /proc/kallsyms | grep ksys_read ffffffff8136d580 T ksys_read # objdump -d --start-address=0xffffffff8136d580 --stop-address=0xffffffff8136d590 /proc/kcore
/proc/kcore: file format elf64-x86-64
Segmentation fault
krava33 login: [ 68.330612] general protection fault, probably for non-canonical address 0xf887ffcbff000: 0000 [#1] SMP PTI [ 68.333118] CPU: 12 PID: 1079 Comm: objdump Not tainted 5.14.0-rc5qemu+ #508 [ 68.334922] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-4.fc34 04/01/2014 [ 68.336945] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.338082] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.342220] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.343428] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.345029] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.346599] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.349000] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.350804] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.352609] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.354638] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.356104] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.357896] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.359694] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.361597] PKRU: 55555554 [ 68.362460] Call Trace: [ 68.363252] read_kcore+0x57f/0x920 [ 68.364289] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.365630] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.366955] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.368277] ? trace_hardirqs_on+0x1b/0xd0 [ 68.369462] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.370793] ? lock_acquire+0x195/0x2f0 [ 68.371920] ? lock_acquire+0x195/0x2f0 [ 68.373035] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.374364] ? lock_acquire+0x195/0x2f0 [ 68.375498] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.376831] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.379883] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.381268] ? lock_release+0x22b/0x3e0 [ 68.382458] ? _raw_spin_unlock+0x1f/0x30 [ 68.383685] ? __handle_mm_fault+0xcfc/0x15f0 [ 68.384994] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.386389] ? lock_acquire+0x195/0x2f0 [ 68.387573] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.388969] ? lock_release+0x22b/0x3e0 [ 68.390145] proc_reg_read+0x55/0xa0 [ 68.391257] ? vfs_read+0x78/0x1b0 [ 68.392336] vfs_read+0xa7/0x1b0 [ 68.393328] ksys_read+0x68/0xe0 [ 68.394308] do_syscall_64+0x3b/0x90 [ 68.395391] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 68.396804] RIP: 0033:0x7fcc11cf92e2 [ 68.397824] Code: c0 e9 b2 fe ff ff 50 48 8d 3d ea 2e 0a 00 e8 95 e9 01 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24 [ 68.402420] RSP: 002b:00007ffd6e0f8da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 [ 68.404357] RAX: ffffffffffffffda RBX: 0000565439305b20 RCX: 00007fcc11cf92e2 [ 68.406061] RDX: 0000000000800000 RSI: 00007fcc0f980010 RDI: 0000000000000003 [ 68.407747] RBP: 00007fcc11dcd300 R08: 0000000000000003 R09: 00007fcc0d980010 [ 68.410937] R10: 0000000003826000 R11: 0000000000000246 R12: 00007fcc0f980010 [ 68.412624] R13: 0000000000000d68 R14: 00007fcc11dcc700 R15: 0000000000800000 [ 68.414322] Modules linked in: intel_rapl_msr intel_rapl_common nfit kvm_intel kvm irqbypass rapl iTCO_wdt iTCO_vendor_support i2c_i801 i2c_smbus lpc_ich drm drm_panel_orientation_quirks zram xfs crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel [ 68.419591] ---[ end trace e2c30f827226966b ]--- [ 68.420969] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.422308] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.426826] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.428150] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.429813] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.431465] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.433115] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.434768] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.436423] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.438354] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.442077] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.443727] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.445370] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.447010] PKRU: 55555554
The fault happens because kern_addr_valid() dereferences existent but not present PMD in the high kernel mappings.
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Make kern_addr_valid() to check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Mike Rapoport rppt@linux.ibm.com Cc: stable@vger.kernel.org # 4.4+ ---
v2: * drop pXd_none() checks and leave only pXd_present(), per David
v1: https://lore.kernel.org/lkml/20210817135854.25407-1-rppt@kernel.org
arch/x86/mm/init_64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ddeaba947eb3..879886c6cc53 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1433,18 +1433,18 @@ int kern_addr_valid(unsigned long addr) return 0;
p4d = p4d_offset(pgd, addr); - if (p4d_none(*p4d)) + if (!p4d_present(*p4d)) return 0;
pud = pud_offset(p4d, addr); - if (pud_none(*pud)) + if (!pud_present(*pud)) return 0;
if (pud_large(*pud)) return pfn_valid(pud_pfn(*pud));
pmd = pmd_offset(pud, addr); - if (pmd_none(*pmd)) + if (!pmd_present(*pmd)) return 0;
if (pmd_large(*pmd))
On 19.08.21 15:27, Mike Rapoport wrote:
From: Mike Rapoport rppt@linux.ibm.com
Jiri Olsa reported a fault when running:
# cat /proc/kallsyms | grep ksys_read ffffffff8136d580 T ksys_read # objdump -d --start-address=0xffffffff8136d580 --stop-address=0xffffffff8136d590 /proc/kcore
/proc/kcore: file format elf64-x86-64
Segmentation fault
krava33 login: [ 68.330612] general protection fault, probably for non-canonical address 0xf887ffcbff000: 0000 [#1] SMP PTI [ 68.333118] CPU: 12 PID: 1079 Comm: objdump Not tainted 5.14.0-rc5qemu+ #508 [ 68.334922] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-4.fc34 04/01/2014 [ 68.336945] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.338082] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.342220] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.343428] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.345029] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.346599] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.349000] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.350804] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.352609] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.354638] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.356104] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.357896] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.359694] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.361597] PKRU: 55555554 [ 68.362460] Call Trace: [ 68.363252] read_kcore+0x57f/0x920 [ 68.364289] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.365630] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.366955] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.368277] ? trace_hardirqs_on+0x1b/0xd0 [ 68.369462] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.370793] ? lock_acquire+0x195/0x2f0 [ 68.371920] ? lock_acquire+0x195/0x2f0 [ 68.373035] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.374364] ? lock_acquire+0x195/0x2f0 [ 68.375498] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.376831] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.379883] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.381268] ? lock_release+0x22b/0x3e0 [ 68.382458] ? _raw_spin_unlock+0x1f/0x30 [ 68.383685] ? __handle_mm_fault+0xcfc/0x15f0 [ 68.384994] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.386389] ? lock_acquire+0x195/0x2f0 [ 68.387573] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.388969] ? lock_release+0x22b/0x3e0 [ 68.390145] proc_reg_read+0x55/0xa0 [ 68.391257] ? vfs_read+0x78/0x1b0 [ 68.392336] vfs_read+0xa7/0x1b0 [ 68.393328] ksys_read+0x68/0xe0 [ 68.394308] do_syscall_64+0x3b/0x90 [ 68.395391] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 68.396804] RIP: 0033:0x7fcc11cf92e2 [ 68.397824] Code: c0 e9 b2 fe ff ff 50 48 8d 3d ea 2e 0a 00 e8 95 e9 01 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24 [ 68.402420] RSP: 002b:00007ffd6e0f8da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 [ 68.404357] RAX: ffffffffffffffda RBX: 0000565439305b20 RCX: 00007fcc11cf92e2 [ 68.406061] RDX: 0000000000800000 RSI: 00007fcc0f980010 RDI: 0000000000000003 [ 68.407747] RBP: 00007fcc11dcd300 R08: 0000000000000003 R09: 00007fcc0d980010 [ 68.410937] R10: 0000000003826000 R11: 0000000000000246 R12: 00007fcc0f980010 [ 68.412624] R13: 0000000000000d68 R14: 00007fcc11dcc700 R15: 0000000000800000 [ 68.414322] Modules linked in: intel_rapl_msr intel_rapl_common nfit kvm_intel kvm irqbypass rapl iTCO_wdt iTCO_vendor_support i2c_i801 i2c_smbus lpc_ich drm drm_panel_orientation_quirks zram xfs crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel [ 68.419591] ---[ end trace e2c30f827226966b ]--- [ 68.420969] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.422308] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.426826] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.428150] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.429813] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.431465] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.433115] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.434768] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.436423] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.438354] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.442077] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.443727] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.445370] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.447010] PKRU: 55555554
The fault happens because kern_addr_valid() dereferences existent but not present PMD in the high kernel mappings.
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Make kern_addr_valid() to check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Mike Rapoport rppt@linux.ibm.com Cc: stable@vger.kernel.org # 4.4+
v2:
- drop pXd_none() checks and leave only pXd_present(), per David
v1: https://lore.kernel.org/lkml/20210817135854.25407-1-rppt@kernel.org
arch/x86/mm/init_64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ddeaba947eb3..879886c6cc53 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1433,18 +1433,18 @@ int kern_addr_valid(unsigned long addr) return 0; p4d = p4d_offset(pgd, addr);
- if (p4d_none(*p4d))
- if (!p4d_present(*p4d)) return 0;
pud = pud_offset(p4d, addr);
- if (pud_none(*pud))
- if (!pud_present(*pud)) return 0;
if (pud_large(*pud)) return pfn_valid(pud_pfn(*pud)); pmd = pmd_offset(pud, addr);
- if (pmd_none(*pmd))
- if (!pmd_present(*pmd)) return 0;
if (pmd_large(*pmd))
Hopefully we won't have other similar BUGs in the code because we leave fake swap entries lying around in the direct map.
Thanks!
Reviewed-by: David Hildenbrand david@redhat.com
On Thu, Aug 19, 2021 at 04:27:17PM +0300, Mike Rapoport wrote:
From: Mike Rapoport rppt@linux.ibm.com
Jiri Olsa reported a fault when running:
# cat /proc/kallsyms | grep ksys_read ffffffff8136d580 T ksys_read # objdump -d --start-address=0xffffffff8136d580 --stop-address=0xffffffff8136d590 /proc/kcore
/proc/kcore: file format elf64-x86-64
Segmentation fault
krava33 login: [ 68.330612] general protection fault, probably for non-canonical address 0xf887ffcbff000: 0000 [#1] SMP PTI [ 68.333118] CPU: 12 PID: 1079 Comm: objdump Not tainted 5.14.0-rc5qemu+ #508 [ 68.334922] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-4.fc34 04/01/2014 [ 68.336945] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.338082] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.342220] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.343428] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.345029] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.346599] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.349000] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.350804] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.352609] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.354638] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.356104] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.357896] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.359694] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.361597] PKRU: 55555554 [ 68.362460] Call Trace: [ 68.363252] read_kcore+0x57f/0x920 [ 68.364289] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.365630] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.366955] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.368277] ? trace_hardirqs_on+0x1b/0xd0 [ 68.369462] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.370793] ? lock_acquire+0x195/0x2f0 [ 68.371920] ? lock_acquire+0x195/0x2f0 [ 68.373035] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.374364] ? lock_acquire+0x195/0x2f0 [ 68.375498] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.376831] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.379883] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.381268] ? lock_release+0x22b/0x3e0 [ 68.382458] ? _raw_spin_unlock+0x1f/0x30 [ 68.383685] ? __handle_mm_fault+0xcfc/0x15f0 [ 68.384994] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.386389] ? lock_acquire+0x195/0x2f0 [ 68.387573] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.388969] ? lock_release+0x22b/0x3e0 [ 68.390145] proc_reg_read+0x55/0xa0 [ 68.391257] ? vfs_read+0x78/0x1b0 [ 68.392336] vfs_read+0xa7/0x1b0 [ 68.393328] ksys_read+0x68/0xe0 [ 68.394308] do_syscall_64+0x3b/0x90 [ 68.395391] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 68.396804] RIP: 0033:0x7fcc11cf92e2 [ 68.397824] Code: c0 e9 b2 fe ff ff 50 48 8d 3d ea 2e 0a 00 e8 95 e9 01 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24 [ 68.402420] RSP: 002b:00007ffd6e0f8da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 [ 68.404357] RAX: ffffffffffffffda RBX: 0000565439305b20 RCX: 00007fcc11cf92e2 [ 68.406061] RDX: 0000000000800000 RSI: 00007fcc0f980010 RDI: 0000000000000003 [ 68.407747] RBP: 00007fcc11dcd300 R08: 0000000000000003 R09: 00007fcc0d980010 [ 68.410937] R10: 0000000003826000 R11: 0000000000000246 R12: 00007fcc0f980010 [ 68.412624] R13: 0000000000000d68 R14: 00007fcc11dcc700 R15: 0000000000800000 [ 68.414322] Modules linked in: intel_rapl_msr intel_rapl_common nfit kvm_intel kvm irqbypass rapl iTCO_wdt iTCO_vendor_support i2c_i801 i2c_smbus lpc_ich drm drm_panel_orientation_quirks zram xfs crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel [ 68.419591] ---[ end trace e2c30f827226966b ]--- [ 68.420969] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.422308] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.426826] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.428150] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.429813] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.431465] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.433115] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.434768] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.436423] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.438354] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.442077] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.443727] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.445370] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.447010] PKRU: 55555554
The fault happens because kern_addr_valid() dereferences existent but not present PMD in the high kernel mappings.
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Make kern_addr_valid() to check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Reported-by: Jiri Olsa jolsa@redhat.com
Tested-by: Jiri Olsa jolsa@redhat.com
thanks, jirka
Signed-off-by: Mike Rapoport rppt@linux.ibm.com Cc: stable@vger.kernel.org # 4.4+
v2:
- drop pXd_none() checks and leave only pXd_present(), per David
v1: https://lore.kernel.org/lkml/20210817135854.25407-1-rppt@kernel.org
arch/x86/mm/init_64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ddeaba947eb3..879886c6cc53 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1433,18 +1433,18 @@ int kern_addr_valid(unsigned long addr) return 0; p4d = p4d_offset(pgd, addr);
- if (p4d_none(*p4d))
- if (!p4d_present(*p4d)) return 0;
pud = pud_offset(p4d, addr);
- if (pud_none(*pud))
- if (!pud_present(*pud)) return 0;
if (pud_large(*pud)) return pfn_valid(pud_pfn(*pud)); pmd = pmd_offset(pud, addr);
- if (pmd_none(*pmd))
- if (!pmd_present(*pmd)) return 0;
if (pmd_large(*pmd)) -- 2.28.0
On 8/19/21 6:27 AM, Mike Rapoport wrote:
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Make kern_addr_valid() to check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Mike Rapoport rppt@linux.ibm.com Cc: stable@vger.kernel.org # 4.4... pmd = pmd_offset(pud, addr);
- if (pmd_none(*pmd))
- if (!pmd_present(*pmd)) return 0;
Yeah, that seems like the right fix. The one kern_addr_valid() user is going to touch the memory so it *better* be present. p*d_none() was definitely the wrong check.
Acked-by: Dave Hansen dave.hansen@intel.com
On Wed, Aug 25, 2021 at 11:47:10AM -0700, Dave Hansen wrote:
On 8/19/21 6:27 AM, Mike Rapoport wrote:
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Make kern_addr_valid() to check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Mike Rapoport rppt@linux.ibm.com Cc: stable@vger.kernel.org # 4.4... pmd = pmd_offset(pud, addr);
- if (pmd_none(*pmd))
- if (!pmd_present(*pmd)) return 0;
Yeah, that seems like the right fix. The one kern_addr_valid() user is going to touch the memory so it *better* be present. p*d_none() was definitely the wrong check.
Acked-by: Dave Hansen dave.hansen@intel.com
So I did stare at this for a while, trying to make sense of it and David Hildenbrand asked for a Fixes: tag in v1 review and from doing a bit of git archeology I think it should be:
c40a56a7818c ("x86/mm/init: Remove freed kernel image areas from alias mapping")
because that thing added the clearing of the Present bit for the high kernel image mapping of those areas.
Right?
On Wed, Sep 08, 2021 at 12:35:21PM +0200, Borislav Petkov wrote:
So I did stare at this for a while, trying to make sense of it and David Hildenbrand asked for a Fixes: tag in v1 review and from doing a bit of git archeology I think it should be:
c40a56a7818c ("x86/mm/init: Remove freed kernel image areas from alias mapping")
because that thing added the clearing of the Present bit for the high kernel image mapping of those areas.
Right?
Hmm, but that commit is in v4.19. Mike has added
Cc: stable@vger.kernel.org # 4.4+
Mike, why 4.4 and newer?
Hmmm.
On Wed, Sep 08, 2021 at 12:52:45PM +0200, Borislav Petkov wrote:
On Wed, Sep 08, 2021 at 12:35:21PM +0200, Borislav Petkov wrote:
So I did stare at this for a while, trying to make sense of it and David Hildenbrand asked for a Fixes: tag in v1 review and from doing a bit of git archeology I think it should be:
c40a56a7818c ("x86/mm/init: Remove freed kernel image areas from alias mapping")
because that thing added the clearing of the Present bit for the high kernel image mapping of those areas.
Right?
Yes, in a sense. As the only user of kern_addr_valid() is kcore and it only uses this check for high kernel mappings, there should be no problem before 4.19.
But...
Hmm, but that commit is in v4.19. Mike has added
Cc: stable@vger.kernel.org # 4.4+
Mike, why 4.4 and newer?
kern_addr_valid() wrongly uses pxy_none() rather than pxy_present() because according to 9a14aefc1d28 ("x86: cpa, fix lookup_address") there could be cases when page table entries exist but they are not valid. So a call to kern_addr_valid() for an address in the direct map would oops.
I've stopped digging at 9a14aefc1d28 (which is in v2.6.26) and added the oldest stable we still support (4.4).
I agree that before 4.19 it's more of a theoretical bug, but you know, things happen...
Hmmm.
On Wed, Sep 08, 2021 at 02:22:31PM +0300, Mike Rapoport wrote:
kern_addr_valid() wrongly uses pxy_none() rather than pxy_present() because according to 9a14aefc1d28 ("x86: cpa, fix lookup_address") there could be cases when page table entries exist but they are not valid. So a call to kern_addr_valid() for an address in the direct map would oops.
I've stopped digging at 9a14aefc1d28 (which is in v2.6.26) and added the oldest stable we still support (4.4).
I agree that before 4.19 it's more of a theoretical bug, but you know, things happen...
Hmmkay, I guess I should add the gist of that to the commit message so that it is explained why 4.4.
I'm assuming the pxy_present() check is more strict than pxy_none() so that backporting to all stable kernels should not introduce any risks...
Any updates on this?
On Thu, Aug 19, 2021 at 04:27:17PM +0300, Mike Rapoport wrote:
From: Mike Rapoport rppt@linux.ibm.com
Jiri Olsa reported a fault when running:
# cat /proc/kallsyms | grep ksys_read ffffffff8136d580 T ksys_read # objdump -d --start-address=0xffffffff8136d580 --stop-address=0xffffffff8136d590 /proc/kcore
/proc/kcore: file format elf64-x86-64
Segmentation fault
krava33 login: [ 68.330612] general protection fault, probably for non-canonical address 0xf887ffcbff000: 0000 [#1] SMP PTI [ 68.333118] CPU: 12 PID: 1079 Comm: objdump Not tainted 5.14.0-rc5qemu+ #508 [ 68.334922] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-4.fc34 04/01/2014 [ 68.336945] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.338082] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.342220] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.343428] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.345029] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.346599] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.349000] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.350804] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.352609] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.354638] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.356104] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.357896] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.359694] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.361597] PKRU: 55555554 [ 68.362460] Call Trace: [ 68.363252] read_kcore+0x57f/0x920 [ 68.364289] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.365630] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.366955] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.368277] ? trace_hardirqs_on+0x1b/0xd0 [ 68.369462] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.370793] ? lock_acquire+0x195/0x2f0 [ 68.371920] ? lock_acquire+0x195/0x2f0 [ 68.373035] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.374364] ? lock_acquire+0x195/0x2f0 [ 68.375498] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.376831] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.379883] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.381268] ? lock_release+0x22b/0x3e0 [ 68.382458] ? _raw_spin_unlock+0x1f/0x30 [ 68.383685] ? __handle_mm_fault+0xcfc/0x15f0 [ 68.384994] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.386389] ? lock_acquire+0x195/0x2f0 [ 68.387573] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.388969] ? lock_release+0x22b/0x3e0 [ 68.390145] proc_reg_read+0x55/0xa0 [ 68.391257] ? vfs_read+0x78/0x1b0 [ 68.392336] vfs_read+0xa7/0x1b0 [ 68.393328] ksys_read+0x68/0xe0 [ 68.394308] do_syscall_64+0x3b/0x90 [ 68.395391] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 68.396804] RIP: 0033:0x7fcc11cf92e2 [ 68.397824] Code: c0 e9 b2 fe ff ff 50 48 8d 3d ea 2e 0a 00 e8 95 e9 01 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24 [ 68.402420] RSP: 002b:00007ffd6e0f8da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 [ 68.404357] RAX: ffffffffffffffda RBX: 0000565439305b20 RCX: 00007fcc11cf92e2 [ 68.406061] RDX: 0000000000800000 RSI: 00007fcc0f980010 RDI: 0000000000000003 [ 68.407747] RBP: 00007fcc11dcd300 R08: 0000000000000003 R09: 00007fcc0d980010 [ 68.410937] R10: 0000000003826000 R11: 0000000000000246 R12: 00007fcc0f980010 [ 68.412624] R13: 0000000000000d68 R14: 00007fcc11dcc700 R15: 0000000000800000 [ 68.414322] Modules linked in: intel_rapl_msr intel_rapl_common nfit kvm_intel kvm irqbypass rapl iTCO_wdt iTCO_vendor_support i2c_i801 i2c_smbus lpc_ich drm drm_panel_orientation_quirks zram xfs crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel [ 68.419591] ---[ end trace e2c30f827226966b ]--- [ 68.420969] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.422308] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.426826] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.428150] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.429813] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.431465] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.433115] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.434768] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.436423] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.438354] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.442077] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.443727] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.445370] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.447010] PKRU: 55555554
The fault happens because kern_addr_valid() dereferences existent but not present PMD in the high kernel mappings.
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Make kern_addr_valid() to check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Mike Rapoport rppt@linux.ibm.com Cc: stable@vger.kernel.org # 4.4+
v2:
- drop pXd_none() checks and leave only pXd_present(), per David
v1: https://lore.kernel.org/lkml/20210817135854.25407-1-rppt@kernel.org
arch/x86/mm/init_64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ddeaba947eb3..879886c6cc53 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1433,18 +1433,18 @@ int kern_addr_valid(unsigned long addr) return 0; p4d = p4d_offset(pgd, addr);
- if (p4d_none(*p4d))
- if (!p4d_present(*p4d)) return 0;
pud = pud_offset(p4d, addr);
- if (pud_none(*pud))
- if (!pud_present(*pud)) return 0;
if (pud_large(*pud)) return pfn_valid(pud_pfn(*pud)); pmd = pmd_offset(pud, addr);
- if (pmd_none(*pmd))
- if (!pmd_present(*pmd)) return 0;
if (pmd_large(*pmd)) -- 2.28.0
Ping?
On Thu, Aug 19, 2021 at 04:27:17PM +0300, Mike Rapoport wrote:
From: Mike Rapoport rppt@linux.ibm.com
Jiri Olsa reported a fault when running:
# cat /proc/kallsyms | grep ksys_read ffffffff8136d580 T ksys_read # objdump -d --start-address=0xffffffff8136d580 --stop-address=0xffffffff8136d590 /proc/kcore
/proc/kcore: file format elf64-x86-64
Segmentation fault
krava33 login: [ 68.330612] general protection fault, probably for non-canonical address 0xf887ffcbff000: 0000 [#1] SMP PTI [ 68.333118] CPU: 12 PID: 1079 Comm: objdump Not tainted 5.14.0-rc5qemu+ #508 [ 68.334922] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-4.fc34 04/01/2014 [ 68.336945] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.338082] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.342220] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.343428] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.345029] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.346599] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.349000] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.350804] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.352609] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.354638] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.356104] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.357896] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.359694] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.361597] PKRU: 55555554 [ 68.362460] Call Trace: [ 68.363252] read_kcore+0x57f/0x920 [ 68.364289] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.365630] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.366955] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.368277] ? trace_hardirqs_on+0x1b/0xd0 [ 68.369462] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.370793] ? lock_acquire+0x195/0x2f0 [ 68.371920] ? lock_acquire+0x195/0x2f0 [ 68.373035] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.374364] ? lock_acquire+0x195/0x2f0 [ 68.375498] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.376831] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.379883] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.381268] ? lock_release+0x22b/0x3e0 [ 68.382458] ? _raw_spin_unlock+0x1f/0x30 [ 68.383685] ? __handle_mm_fault+0xcfc/0x15f0 [ 68.384994] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.386389] ? lock_acquire+0x195/0x2f0 [ 68.387573] ? rcu_read_lock_sched_held+0x12/0x80 [ 68.388969] ? lock_release+0x22b/0x3e0 [ 68.390145] proc_reg_read+0x55/0xa0 [ 68.391257] ? vfs_read+0x78/0x1b0 [ 68.392336] vfs_read+0xa7/0x1b0 [ 68.393328] ksys_read+0x68/0xe0 [ 68.394308] do_syscall_64+0x3b/0x90 [ 68.395391] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 68.396804] RIP: 0033:0x7fcc11cf92e2 [ 68.397824] Code: c0 e9 b2 fe ff ff 50 48 8d 3d ea 2e 0a 00 e8 95 e9 01 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24 [ 68.402420] RSP: 002b:00007ffd6e0f8da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 [ 68.404357] RAX: ffffffffffffffda RBX: 0000565439305b20 RCX: 00007fcc11cf92e2 [ 68.406061] RDX: 0000000000800000 RSI: 00007fcc0f980010 RDI: 0000000000000003 [ 68.407747] RBP: 00007fcc11dcd300 R08: 0000000000000003 R09: 00007fcc0d980010 [ 68.410937] R10: 0000000003826000 R11: 0000000000000246 R12: 00007fcc0f980010 [ 68.412624] R13: 0000000000000d68 R14: 00007fcc11dcc700 R15: 0000000000800000 [ 68.414322] Modules linked in: intel_rapl_msr intel_rapl_common nfit kvm_intel kvm irqbypass rapl iTCO_wdt iTCO_vendor_support i2c_i801 i2c_smbus lpc_ich drm drm_panel_orientation_quirks zram xfs crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel [ 68.419591] ---[ end trace e2c30f827226966b ]--- [ 68.420969] RIP: 0010:kern_addr_valid+0x150/0x300 [ 68.422308] Code: 1f 40 00 48 8b 0d e8 12 61 01 48 85 f6 0f 85 ca 00 00 00 48 81 e1 00 f0 ff ff 48 21 c1 48 b8 00 00 00 00 80 88 ff ff 48 01 ca <48> 8b 3c 02 48 f7 c7 9f ff ff ff 0f 84 d8 fe ff ff 48 89 f8 0f 1f [ 68.426826] RSP: 0018:ffffc90000bcbc38 EFLAGS: 00010206 [ 68.428150] RAX: ffff888000000000 RBX: 0000000000001000 RCX: 000ffffffcbff000 [ 68.429813] RDX: 000ffffffcbff000 RSI: 0000000000000000 RDI: 800ffffffcbff062 [ 68.431465] RBP: ffffc90000bcbea8 R08: 0000000000001000 R09: 0000000000000000 [ 68.433115] R10: 0000000000000000 R11: 0000000000001000 R12: 00007fcc0fd80010 [ 68.434768] R13: ffffffff83400000 R14: 0000000000400000 R15: ffffffff843d23e0 [ 68.436423] FS: 00007fcc111fcc80(0000) GS:ffff888275e00000(0000) knlGS:0000000000000000 [ 68.438354] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 68.442077] CR2: 00007fcc0fd80000 CR3: 000000011226e004 CR4: 0000000000770ee0 [ 68.443727] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 68.445370] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 68.447010] PKRU: 55555554
The fault happens because kern_addr_valid() dereferences existent but not present PMD in the high kernel mappings.
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Make kern_addr_valid() to check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Mike Rapoport rppt@linux.ibm.com Cc: stable@vger.kernel.org # 4.4+
v2:
- drop pXd_none() checks and leave only pXd_present(), per David
v1: https://lore.kernel.org/lkml/20210817135854.25407-1-rppt@kernel.org
arch/x86/mm/init_64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ddeaba947eb3..879886c6cc53 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1433,18 +1433,18 @@ int kern_addr_valid(unsigned long addr) return 0; p4d = p4d_offset(pgd, addr);
- if (p4d_none(*p4d))
- if (!p4d_present(*p4d)) return 0;
pud = pud_offset(p4d, addr);
- if (pud_none(*pud))
- if (!pud_present(*pud)) return 0;
if (pud_large(*pud)) return pfn_valid(pud_pfn(*pud)); pmd = pmd_offset(pud, addr);
- if (pmd_none(*pmd))
- if (!pmd_present(*pmd)) return 0;
if (pmd_large(*pmd)) -- 2.28.0
The following commit has been merged into the x86/urgent branch of tip:
Commit-ID: 34b1999da935a33be6239226bfa6cd4f704c5c88 Gitweb: https://git.kernel.org/tip/34b1999da935a33be6239226bfa6cd4f704c5c88 Author: Mike Rapoport rppt@linux.ibm.com AuthorDate: Thu, 19 Aug 2021 16:27:17 +03:00 Committer: Borislav Petkov bp@suse.de CommitterDate: Wed, 08 Sep 2021 20:50:32 +02:00
x86/mm: Fix kern_addr_valid() to cope with existing but not present entries
Jiri Olsa reported a fault when running:
# cat /proc/kallsyms | grep ksys_read ffffffff8136d580 T ksys_read # objdump -d --start-address=0xffffffff8136d580 --stop-address=0xffffffff8136d590 /proc/kcore
/proc/kcore: file format elf64-x86-64
Segmentation fault
general protection fault, probably for non-canonical address 0xf887ffcbff000: 0000 [#1] SMP PTI CPU: 12 PID: 1079 Comm: objdump Not tainted 5.14.0-rc5qemu+ #508 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-4.fc34 04/01/2014 RIP: 0010:kern_addr_valid Call Trace: read_kcore ? rcu_read_lock_sched_held ? rcu_read_lock_sched_held ? rcu_read_lock_sched_held ? trace_hardirqs_on ? rcu_read_lock_sched_held ? lock_acquire ? lock_acquire ? rcu_read_lock_sched_held ? lock_acquire ? rcu_read_lock_sched_held ? rcu_read_lock_sched_held ? rcu_read_lock_sched_held ? lock_release ? _raw_spin_unlock ? __handle_mm_fault ? rcu_read_lock_sched_held ? lock_acquire ? rcu_read_lock_sched_held ? lock_release proc_reg_read ? vfs_read vfs_read ksys_read do_syscall_64 entry_SYSCALL_64_after_hwframe
The fault happens because kern_addr_valid() dereferences existent but not present PMD in the high kernel mappings.
Such PMDs are created when free_kernel_image_pages() frees regions larger than 2Mb. In this case, a part of the freed memory is mapped with PMDs and the set_memory_np_noalias() -> ... -> __change_page_attr() sequence will mark the PMD as not present rather than wipe it completely.
Have kern_addr_valid() check whether higher level page table entries are present before trying to dereference them to fix this issue and to avoid similar issues in the future.
Stable backporting note: ------------------------
Note that the stable marking is for all active stable branches because there could be cases where pagetable entries exist but are not valid - see 9a14aefc1d28 ("x86: cpa, fix lookup_address"), for example. So make sure to be on the safe side here and use pXY_present() accessors rather than pXY_none() which could #GP when accessing pages in the direct map.
Also see:
c40a56a7818c ("x86/mm/init: Remove freed kernel image areas from alias mapping")
for more info.
Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Mike Rapoport rppt@linux.ibm.com Signed-off-by: Borislav Petkov bp@suse.de Reviewed-by: David Hildenbrand david@redhat.com Acked-by: Dave Hansen dave.hansen@intel.com Tested-by: Jiri Olsa jolsa@redhat.com Cc: stable@vger.kernel.org # 4.4+ Link: https://lkml.kernel.org/r/20210819132717.19358-1-rppt@kernel.org --- arch/x86/mm/init_64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ddeaba9..879886c 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1433,18 +1433,18 @@ int kern_addr_valid(unsigned long addr) return 0;
p4d = p4d_offset(pgd, addr); - if (p4d_none(*p4d)) + if (!p4d_present(*p4d)) return 0;
pud = pud_offset(p4d, addr); - if (pud_none(*pud)) + if (!pud_present(*pud)) return 0;
if (pud_large(*pud)) return pfn_valid(pud_pfn(*pud));
pmd = pmd_offset(pud, addr); - if (pmd_none(*pmd)) + if (!pmd_present(*pmd)) return 0;
if (pmd_large(*pmd))
linux-stable-mirror@lists.linaro.org
-
Borislav Petkov -
Dave Hansen -
David Hildenbrand -
Jiri Olsa -
Mike Rapoport -
tip-bot2 for Mike Rapoport