AuxXxilium/linux_dsm_epyc7002
1
0
Fork 0
mirror of https://github.com/AuxXxilium/linux_dsm_epyc7002.git synced 2024-11-29 20:06:38 +07:00
Code Issues Actions Packages Projects Releases Wiki Activity

powerpc/eeh: Fix kernel crash when passing through VF

Browse Source 
When doing vfio passthrough a VF, the kernel will crash with following
message:

[  442.656459] Unable to handle kernel paging request for data at address 0x00000060
[  442.656593] Faulting instruction address: 0xc000000000038b88
[  442.656706] Oops: Kernel access of bad area, sig: 11 [#1]
[  442.656798] SMP NR_CPUS=1024 NUMA PowerNV
[  442.656890] Modules linked in: vfio_pci mlx4_core nf_conntrack_netbios_ns nf_conntrack_broadcast ipt_MASQUERADE ip6t_REJECT xt_conntrack bnep bluetooth rfkill ebtable_nat ebtable_broute bridge stp llc ebtable_filter ebtables ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw ip6table_filter ip6_tables iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_mangle iptable_security iptable_raw tg3 nfsd be2net nfs_acl ses lockd ptp enclosure pps_core kvm_hv kvm_pr shpchp binfmt_misc kvm sunrpc uinput lpfc scsi_transport_fc ipr scsi_tgt [last unloaded: mlx4_core]
[  442.658152] CPU: 40 PID: 14948 Comm: qemu-system-ppc Not tainted 3.10.42yw-pkvm+ #37
[  442.658219] task: c000000f7e2a9a00 ti: c000000f6dc3c000 task.ti: c000000f6dc3c000
[  442.658287] NIP: c000000000038b88 LR: c0000000004435a8 CTR: c000000000455bc0
[  442.658352] REGS: c000000f6dc3f580 TRAP: 0300   Not tainted  (3.10.42yw-pkvm+)
[  442.658419] MSR: 9000000000009032 <SF,HV,EE,ME,IR,DR,RI>  CR: 28004882  XER: 20000000
[  442.658577] CFAR: c00000000000908c DAR: 0000000000000060 DSISR: 40000000 SOFTE: 1
GPR00: c0000000004435a8 c000000f6dc3f800 c0000000012b1c10 c00000000da24000
GPR04: 0000000000000003 0000000000001004 00000000000015b3 000000000000ffff
GPR08: c00000000127f5d8 0000000000000000 000000000000ffff 0000000000000000
GPR12: c000000000068078 c00000000fdd6800 000001003c320c80 000001003c3607f0
GPR16: 0000000000000001 00000000105480c8 000000001055aaa8 000001003c31ab18
GPR20: 000001003c10fb40 000001003c360ae8 000000001063bcf0 000000001063bdb0
GPR24: 000001003c15ed70 0000000010548f40 c000001fe5514c88 c000001fe5514cb0
GPR28: c00000000da24000 0000000000000000 c00000000da24000 0000000000000003
[  442.659471] NIP [c000000000038b88] .pcibios_set_pcie_reset_state+0x28/0x130
[  442.659530] LR [c0000000004435a8] .pci_set_pcie_reset_state+0x28/0x40
[  442.659585] Call Trace:
[  442.659610] [c000000f6dc3f800] [00000000000719e0] 0x719e0 (unreliable)
[  442.659677] [c000000f6dc3f880] [c0000000004435a8] .pci_set_pcie_reset_state+0x28/0x40
[  442.659757] [c000000f6dc3f900] [c000000000455bf8] .reset_fundamental+0x38/0x80
[  442.659835] [c000000f6dc3f980] [c0000000004562a8] .pci_dev_specific_reset+0xa8/0xf0
[  442.659913] [c000000f6dc3fa00] [c0000000004448c4] .__pci_dev_reset+0x44/0x430
[  442.659980] [c000000f6dc3fab0] [c000000000444d5c] .pci_reset_function+0x7c/0xc0
[  442.660059] [c000000f6dc3fb30] [d00000001c141ab8] .vfio_pci_open+0xe8/0x2b0 [vfio_pci]
[  442.660139] [c000000f6dc3fbd0] [c000000000586c30] .vfio_group_fops_unl_ioctl+0x3a0/0x630
[  442.660219] [c000000f6dc3fc90] [c000000000255fbc] .do_vfs_ioctl+0x4ec/0x7c0
[  442.660286] [c000000f6dc3fd80] [c000000000256364] .SyS_ioctl+0xd4/0xf0
[  442.660354] [c000000f6dc3fe30] [c000000000009e54] syscall_exit+0x0/0x98
[  442.660420] Instruction dump:
[  442.660454] 4bfffce9 4bfffee4 7c0802a6 fbc1fff0 fbe1fff8 f8010010 f821ff81 7c7e1b78
[  442.660566] 7c9f2378 60000000 60000000 e93e02c8 <e8690060> 2fa30000 41de00c4 2b9f0002
[  442.660679] ---[ end trace a64ac9546bcf0328 ]---
[  442.660724]

The reason is current VF is not EEH enabled.

This patch introduces a macro to convert eeh_dev to eeh_pe. By doing so, it
will prevent converting with NULL pointer.

Signed-off-by: Wei Yang <weiyang@linux.vnet.ibm.com>
Acked-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
CC: Michael Ellerman <mpe@ellerman.id.au>

V3 -> V4:
   1. move the macro definition from include/linux/pci.h to
      arch/powerpc/include/asm/eeh.h

V2 -> V3:
   1. rebased on 3.17-rc4
   2. introduce a macro
   3. use this macro in several other places

V1 -> V2:
   1. code style and patch subject adjustment

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This commit is contained in:
Wei Yang 2014-09-17 10:48:26 +08:00 committed by Michael Ellerman
parent 9e34992a62
commit 2a58222f82
3 changed files with 8 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
arch/powerpc/include/asm/eeh.h
View File

@ -146,6 +146,11 @@ static inline struct pci_dev *eeh_dev_to_pci_dev(struct eeh_dev *edev)
return edev ? edev->pdev : NULL;
}
static inline struct eeh_pe *eeh_dev_to_pe(struct eeh_dev* edev)
{
return edev ? edev->pe : NULL;
}
/* Return values from eeh_ops::next_error */
enum {
EEH_NEXT_ERR_NONE = 0,

4
arch/powerpc/kernel/eeh.c
View File

@ -410,7 +410,7 @@ int eeh_dev_check_failure(struct eeh_dev *edev)
}
dn = eeh_dev_to_of_node(edev);
dev = eeh_dev_to_pci_dev(edev);
pe = edev->pe;
pe = eeh_dev_to_pe(edev);
/* Access to IO BARs might get this far and still not want checking. */
if (!pe) {
@ -634,7 +634,7 @@ int eeh_pci_enable(struct eeh_pe *pe, int function)
int pcibios_set_pcie_reset_state(struct pci_dev *dev, enum pcie_reset_state state)
{
struct eeh_dev *edev = pci_dev_to_eeh_dev(dev);
struct eeh_pe *pe = edev->pe;
struct eeh_pe *pe = eeh_dev_to_pe(edev);
if (!pe) {
pr_err("%s: No PE found on PCI device %s\n",

2
arch/powerpc/kernel/eeh_pe.c
View File

@ -428,7 +428,7 @@ int eeh_rmv_from_parent_pe(struct eeh_dev *edev)
}
/* Remove the EEH device */
pe = edev->pe;
pe = eeh_dev_to_pe(edev);
edev->pe = NULL;
list_del(&edev->list);