Linux/AM3351: DDR self-refresh issues

Aniket Jesu

Part Number: AM3351
Other Parts Discussed in Thread: TPS65217

Tool/software: Linux

Hello TI,

We are having a custom board with TI processor AM3351, when we put processesor in low power mode and put DDR in self-auto refresh mode by changing
emif register 0x4C000038 to this value 0x2A0, we are seeing random kernel/memory exceptions. Can you suggest on this ?

Short Log:

Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_MCM] Waiting for BLE RX - Sem
Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_GAT] Mobile Data Send: OK
Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_GAT]
Jan 16 11:45:14 Mighty-B6 server[207]: [THR_MON] Mighty Heart Beat
Jan 16 11:45:44 Mighty-B6 server[207]: [THR_MON] Mighty Heart Beat
Jan 16 11:45:48 Mighty-B6 server[207]: [MTY_SYS] Exit wait state
Jan 16 11:45:48 Mighty-B6 server[207]: [SON_DWL] Spotify Prefetch Wait Time out
Jan 16 11:45:48 Mighty-B6 server[207]: [SON_DWL] Updating Playlist : 90s Smash Hits : DB_PLAYLIST_OFFLINE_ERROR
Jan 16 11:45:48 Mighty-B6 kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000000
Jan 16 11:45:48 Mighty-B6 kernel: pgd = cc7d8000
Jan 16 11:45:48 Mighty-B6 kernel: [00000000] *pgd=8c5d1831, *pte=00000000, *ppte=00000000
Jan 16 11:45:48 Mighty-B6 kernel: Internal error: Oops: 817 [#1] PREEMPT ARM
Jan 16 11:45:48 Mighty-B6 kernel: Modules linked in: brcmfmac cfg80211 brcmutil
Jan 16 11:45:48 Mighty-B6 kernel: CPU: 0 PID: 247 Comm: server Not tainted 4.1.18-gbbe8cfc #1
Jan 16 11:45:48 Mighty-B6 kernel: Hardware name: Generic AM33XX (Flattened Device Tree)
Jan 16 11:45:48 Mighty-B6 kernel: task: cdde0000 ti: cde52000 task.ti: cde52000
Jan 16 11:45:48 Mighty-B6 kernel: PC is at __kstrtab_netdev_set_default_ethtool_ops+0x13/0x1f
Jan 16 11:45:48 Mighty-B6 kernel: LR is at __kstrtab_netdev_set_default_ethtool_ops+0x13/0x1f
Jan 16 11:45:48 Mighty-B6 kernel: pc : [<c08891f8>] lr : [<c08891f8>] psr: 600f0013
sp : cde53e08 ip : cde53d90 fp : cde53ec4
Jan 16 11:45:48 Mighty-B6 kernel: r10: 000d0d18 r9 : 00000001 r8 : 00000000
Jan 16 11:45:48 Mighty-B6 kernel: r7 : 00000000 r6 : cde52000 r5 : cde53e28 r4 : cde52000
Jan 16 11:45:48 Mighty-B6 kernel: r3 : 00000000 r2 : cde52000 r1 : cdde0000 r0 : fffffffc
Jan 16 11:45:48 Mighty-B6 kernel: Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
Jan 16 11:45:48 Mighty-B6 kernel: Control: 10c5387d Table: 8c7d8019 DAC: 00000015
Jan 16 11:45:48 Mighty-B6 kernel: Process server (pid: 247, stack limit = 0xcde52210)

Full Log:
log_lpm_crash.txt

thanks,
aniket

Fullscreen log_lpm_crash.txt Download

Jan 16 11:43:59 Mighty-B6 server[207]: [BLE_GAT] ** Recvd Mobile Data **
Jan 16 11:43:59 Mighty-B6 server[207]: [MSG_PAC] UnPack MSG  Type: 100  ID: 5
Jan 16 11:43:59 Mighty-B6 server[207]: [THR_IPC] Found Service Hanler TableIndex: 10 ServiceID: 8 HandlerID: 0
Jan 16 11:43:59 Mighty-B6 server[207]: [MSG_PAC] MSG Type: 202, MSG ID: 5
Jan 16 11:43:59 Mighty-B6 server[207]: [MSG_PAC] Pack Bat Capacity: 87% (280mA) Charger Status: 1
Jan 16 11:43:59 Mighty-B6 server[207]: [BLE_MCM] Waiting for BLE RX - Q
Jan 16 11:43:59 Mighty-B6 server[207]: [BLE_MCM] Waiting for BLE RX - Sem
Jan 16 11:43:59 Mighty-B6 server[207]: [BLE_GAT] Mobile Data Send: OK
Jan 16 11:43:59 Mighty-B6 server[207]: [BLE_GAT]
Jan 16 11:44:14 Mighty-B6 server[207]: [THR_MON] Mighty Heart Beat
Jan 16 11:44:44 Mighty-B6 server[207]: [THR_MON] Mighty Heart Beat
Jan 16 11:44:44 Mighty-B6 server[207]: [BAT_MON] BAT driver event, ignored [0]
Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_GAT] ** Recvd Mobile Data **
Jan 16 11:44:59 Mighty-B6 server[207]: [MSG_PAC] UnPack MSG  Type: 100  ID: 5
Jan 16 11:44:59 Mighty-B6 server[207]: [THR_IPC] Found Service Hanler TableIndex: 10 ServiceID: 8 HandlerID: 0
Jan 16 11:44:59 Mighty-B6 server[207]: [MSG_PAC] MSG Type: 202, MSG ID: 5
Jan 16 11:44:59 Mighty-B6 server[207]: [MSG_PAC] Pack Bat Capacity: 85% (280mA) Charger Status: 1
Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_MCM] Waiting for BLE RX - Q
Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_MCM] Waiting for BLE RX - Sem
Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_GAT] Mobile Data Send: OK
Jan 16 11:44:59 Mighty-B6 server[207]: [BLE_GAT]
Jan 16 11:45:14 Mighty-B6 server[207]: [THR_MON] Mighty Heart Beat
Jan 16 11:45:44 Mighty-B6 server[207]: [THR_MON] Mighty Heart Beat
Jan 16 11:45:48 Mighty-B6 server[207]: [MTY_SYS] Exit wait state
Jan 16 11:45:48 Mighty-B6 server[207]: [SON_DWL] Spotify Prefetch Wait Time out
Jan 16 11:45:48 Mighty-B6 server[207]: [SON_DWL] Updating Playlist : 90s Smash Hits : DB_PLAYLIST_OFFLINE_ERROR
Jan 16 11:45:48 Mighty-B6 kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000000
Jan 16 11:45:48 Mighty-B6 kernel: pgd = cc7d8000
Jan 16 11:45:48 Mighty-B6 kernel: [00000000] *pgd=8c5d1831, *pte=00000000, *ppte=00000000
Jan 16 11:45:48 Mighty-B6 kernel: Internal error: Oops: 817 [#1] PREEMPT ARM
Jan 16 11:45:48 Mighty-B6 kernel: Modules linked in: brcmfmac cfg80211 brcmutil
Jan 16 11:45:48 Mighty-B6 kernel: CPU: 0 PID: 247 Comm: server Not tainted 4.1.18-gbbe8cfc #1
Jan 16 11:45:48 Mighty-B6 kernel: Hardware name: Generic AM33XX (Flattened Device Tree)
Jan 16 11:45:48 Mighty-B6 kernel: task: cdde0000 ti: cde52000 task.ti: cde52000
Jan 16 11:45:48 Mighty-B6 kernel: PC is at __kstrtab_netdev_set_default_ethtool_ops+0x13/0x1f
Jan 16 11:45:48 Mighty-B6 kernel: LR is at __kstrtab_netdev_set_default_ethtool_ops+0x13/0x1f
Jan 16 11:45:48 Mighty-B6 kernel: pc : [<c08891f8>]    lr : [<c08891f8>]    psr: 600f0013
                                  sp : cde53e08  ip : cde53d90  fp : cde53ec4
Jan 16 11:45:48 Mighty-B6 kernel: r10: 000d0d18  r9 : 00000001  r8 : 00000000
Jan 16 11:45:48 Mighty-B6 kernel: r7 : 00000000  r6 : cde52000  r5 : cde53e28  r4 : cde52000
Jan 16 11:45:48 Mighty-B6 kernel: r3 : 00000000  r2 : cde52000  r1 : cdde0000  r0 : fffffffc
Jan 16 11:45:48 Mighty-B6 kernel: Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
Jan 16 11:45:48 Mighty-B6 kernel: Control: 10c5387d  Table: 8c7d8019  DAC: 00000015
Jan 16 11:45:48 Mighty-B6 kernel: Process server (pid: 247, stack limit = 0xcde52210)
Jan 16 11:45:48 Mighty-B6 kernel: Stack: (0xcde53e08 to 0xcde54000)
Jan 16 11:45:48 Mighty-B6 kernel: 3e00:                   cde53e24 00004000 00000008 cde53f68 f6906a9a 00004086
Jan 16 11:45:48 Mighty-B6 kernel: 3e20: cde53e38 cd98c1a4 cc235b41 00000000 00000000 00000000 f6106a9a 00004086
Jan 16 11:45:48 Mighty-B6 kernel: 3e40: f60fa74a 00000086 c0077edc c08e1750 00000001 00000000 cdde0000 00000000
Jan 16 11:45:48 Mighty-B6 kernel: 3e60: 00000064 cde53e64 cde53e64 cde53e6c cde53e6c cdde0000 00000000 000d0000
Jan 16 11:45:48 Mighty-B6 kernel: 3e80: cdc8e700 00000d18 00000000 00000000 00000000 ffffffff c0519750 00000000

Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.778936] Internal error: Oops: 817 [#1] PREEMPT ARM

Jan 16 11:45:49 Mighty-B6 kernel: 3ea0: 00000000 00000001 000d0d18 00e4bddc 00000000 000d0d18 cde53f4c cde53ec8
Jan 16 11:45:49 Mighty-B6 kernel: 3ec0: c008acd4 c00890f4 ffffffff 00001000 cde53eb4 00000001 00000000 00000000
Jan 16 11:45:49 Mighty-B6 kernel: 3ee0: c007e288 00000000 3d21b530 00000003 bf80e53a 002081b9 ffffffff 00000000
Jan 16 11:45:49 Mighty-B6 kernel: 3f00: cde53f4c cde53f10 c007e08c c0023148 3d21b530 00000003 bf80e53a 002081b9
Jan 16 11:45:49 Mighty-B6 kernel: 3f20: ec0b0bea 00000000 00000000 00000080 000d0d18 00e4bddc 00000000 00000001
Jan 16 11:45:49 Mighty-B6 kernel: 3f40: cde53fa4 cde53f50 c008b784 c008abe4 b55cb8d8 00000000 ffffffff c007de04
Jan 16 11:45:49 Mighty-B6 kernel: 3f60: f60fa74a 00000086 f60fa74a 00000086 00000000 00e4bddc 00000008 b55cb8d8
Jan 16 11:45:49 Mighty-B6 kernel: 3f80: 000d0d18 00000000 000000f0 c000fa24 cde52000 00000000 00000000 cde53fa8
Jan 16 11:45:49 Mighty-B6 kernel: 3fa0: c000f880 c008b710 b55cb8d8 000d0d18 000d0d18 00000080 00000001 b55cb858
Jan 16 11:45:49 Mighty-B6 kernel: 3fc0: b55cb8d8 000d0d18 00000000 000000f0 00000000 00000000 44b82fa1 3b9aca00
Jan 16 11:45:49 Mighty-B6 kernel: 3fe0: 000000f0 b55cb848 b6e55125 b6e57ca6 20030030 000d0d18 00000000 00000000
Jan 16 11:45:49 Mighty-B6 kernel: Backtrace:
Jan 16 11:45:49 Mighty-B6 kernel: [<c00890e8>] (futex_wait) from [<c008acd4>] (do_futex+0xfc/0xb2c)
Jan 16 11:45:49 Mighty-B6 kernel:  r10:000d0d18 r9:00000000 r8:00e4bddc r7:000d0d18 r6:00000001 r5:00000000
Jan 16 11:45:49 Mighty-B6 kernel:  r4:00000000
Jan 16 11:45:49 Mighty-B6 kernel: [<c008abd8>] (do_futex) from [<c008b784>] (SyS_futex+0x80/0x164)
Jan 16 11:45:49 Mighty-B6 kernel:  r10:00000001 r9:00000000 r8:00e4bddc r7:000d0d18 r6:00000080 r5:00000000
Jan 16 11:45:49 Mighty-B6 kernel:  r4:00000000
Jan 16 11:45:49 Mighty-B6 kernel: [<c008b704>] (SyS_futex) from [<c000f880>] (ret_fast_syscall+0x0/0x3c)
Jan 16 11:45:49 Mighty-B6 kernel:  r10:00000000 r9:cde52000 r8:c000fa24 r7:000000f0 r6:00000000 r5:000d0d18
Jan 16 11:45:49 Mighty-B6 kernel:  r4:b55cb8d8
Jan 16 11:45:49 Mighty-B6 kernel: Code: 5f766564 5f746573 61666564 5f746c75 (74687465)
Jan 16 11:45:49 Mighty-B6 kernel: ---[ end trace da7cb998ff81836b ]---
Jan 16 11:45:48 Mighty-B6 server[207]: [MTY_SYS] SET REFETCH ACTION : 3
Jan 16 11:45:48 Mighty-B6 server[207]: [MTY_SYS] SET REFETCH ACTION : 0
Jan 16 11:45:48 Mighty-B6 server[207]: [SPT_SER] Stopping Prefetch
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.759172] Unable to handle kernel NULL pointer dereference at virtual address 00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.768115] pgd = cc7d8000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.770922] [00000000] *pgd=8c5d1831, *pte=00000000, *ppte=00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.778936] Internal error: Oops: 817 [#1] PREEMPT ARM
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.784164] Modules linked in: brcmfmac cfg80211 brcmutil
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.789816] CPU: 0 PID: 247 Comm: server Not tainted 4.1.18-gbbe8cfc #1
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.796520] Hardware name: Generic AM33XX (Flattened Device Tree)
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.802711] task: cdde0000 ti: cde52000 task.ti: cde52000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.808265] PC is at __kstrtab_netdev_set_default_ethtool_ops+0x13/0x1f
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.814991] LR is at __kstrtab_netdev_set_default_ethtool_ops+0x13/0x1f
Jan 16 11:45:48 Mighty-B6 server[207]: [SON_DWL] Waiting For Spotify Stop Prefetch Callback
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.821715] pc : [<c08891f8>]    lr : [<c08891f8>]    psr: 600f0013
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.821715] sp : cde53e08  ip : cde53d90  fp : cde53ec4

Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.864953] Process server (pid: 247, stack limit = 0xcde52210)


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.870961] Stack: (0xcde53e08 to 0xcde54000)


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.875431] 3e00:                   cde53e24 00004000 00000008 cde53f68 f6906a9a 00004086


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.883745] 3e20: cde53e38 cd98c1a4 cc235b41 00000000 00000000 00000000 f6106a9a 00004086


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.892058] 3e40: f60fa74a 00000086 c0077edc c08e1750 00000001 00000000 cdde0000 00000000


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.900373] 3e60: 00000064 cde53e64 cde53e64 cde53e6c cde53e6c cdde0000 00000000 000d0000

Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.833323] r10: 000d0d18  r9 : 00000001  r8 : 00000000

Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:49 UTC):

kernel[198]: [  878.908684] 3e80: cdc8e700 00000d18 00000000 00000000 00000000 ffffffff c0519750 00000000

Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.838633] r7 : 00000000  r6 : cde52000  r5 : cde53e28  r4 : cde52000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.845253] r3 : 00000000  r2 : cde52000  r1 : cdde0000  r0 : fffffffc
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.851883] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.859117] Control: 10c5387d  Table: 8c7d8019  DAC: 00000015
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.864953] Process server (pid: 247, stack limit = 0xcde52210)
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.870961] Stack: (0xcde53e08 to 0xcde54000)
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.875431] 3e00:                   cde53e24 00004000 00000008 cde53f68 f6906a9a 00004086

Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.916996] 3ea0: 00000000 00000001 000d0d18 00e4bddc 00000000 000d0d18 cde53f4c cde53ec8

Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.883745] 3e20: cde53e38 cd98c1a4 cc235b41 00000000 00000000 00000000 f6106a9a 00004086
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.892058] 3e40: f60fa74a 00000086 c0077edc c08e1750 00000001 00000000 cdde0000 00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.900373] 3e60: 00000064 cde53e64 cde53e64 cde53e6c cde53e6c cdde0000 00000000 000d0000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.908684] 3e80: cdc8e700 00000d18 00000000 00000000 00000000 ffffffff c0519750 00000000

Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.925307] 3ec0: c008acd4 c00890f4 ffffffff 00001000 cde53eb4 00000001 00000000 00000000

Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.916996] 3ea0: 00000000 00000001 000d0d18 00e4bddc 00000000 000d0d18 cde53f4c cde53ec8
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.925307] 3ec0: c008acd4 c00890f4 ffffffff 00001000 cde53eb4 00000001 00000000 00000000

Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.933620] 3ee0: c007e288 00000000 3d21b530 00000003 bf80e53a 002081b9 ffffffff 00000000


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.941933] 3f00: cde53f4c cde53f10 c007e08c c0023148 3d21b530 00000003 bf80e53a 002081b9


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.950244] 3f20: ec0b0bea 00000000 00000000 00000080 000d0d18 00e4bddc 00000000 00000001


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.958558] 3f40: cde53fa4 cde53f50 c008b784 c008abe4 b55cb8d8 00000000 ffffffff c007de04


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.966872] 3f60: f60fa74a 00000086 f60fa74a 00000086 00000000 00e4bddc 00000008 b55cb8d8


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.975185] 3f80: 000d0d18 00000000 000000f0 c000fa24 cde52000 00000000 00000000 cde53fa8


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.983498] 3fa0: c000f880 c008b710 b55cb8d8 000d0d18 000d0d18 00000080 00000001 b55cb858


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  878.991808] 3fc0: b55cb8d8 000d0d18 00000000 000000f0 00000000 00000000 44b82fa1 3b9aca00


Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  879.000124] 3fe0: 000000f0 b55cb848 b6e55125 b6e57ca6 20030030 000d0d18 00000000 00000000

Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.933620] 3ee0: c007e288 00000000 3d21b530 00000003 bf80e53a 002081b9 ffffffff 00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.941933] 3f00: cde53f4c cde53f10 c007e08c c0023148 3d21b530 00000003 bf80e53a 002081b9
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.950244] 3f20: ec0b0bea 00000000 00000000 00000080 000d0d18 00e4bddc 00000000 00000001
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.958558] 3f40: cde53fa4 cde53f50 c008b784 c008abe4 b55cb8d8 00000000 ffffffff c007de04
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.966872] 3f60: f60fa74a 00000086 f60fa74a 00000086 00000000 00e4bddc 00000008 b55cb8d8
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.975185] 3f80: 000d0d18 00000000 000000f0 c000fa24 cde52000 00000000 00000000 cde53fa8
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.983498] 3fa0: c000f880 c008b710 b55cb8d8 000d0d18 000d0d18 00000080 00000001 b55cb858
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  878.991808] 3fc0: b55cb8d8 000d0d18 00000000 000000f0 00000000 00000000 44b82fa1 3b9aca00
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.000124] 3fe0: 000000f0 b55cb848 b6e55125 b6e57ca6 20030030 000d0d18 00000000 00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.008381] Backtrace:
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.011008] [<c00890e8>] (futex_wait) from [<c008acd4>] (do_futex+0xfc/0xb2c)
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.018234]  r10:000d0d18 r9:00000000 r8:00e4bddc r7:000d0d18 r6:00000001 r5:00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.026319]  r4:00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.028993] [<c008abd8>] (do_futex) from [<c008b784>] (SyS_futex+0x80/0x164)

Broadcast message from systemd-journald@Mighty-B6 (Tue 2018-01-16 11:45:50 UTC):

kernel[198]: [  879.065300] Code: 5f766564 5f746573 61666564 5f746c75 (74687465)

Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.036127]  r10:00000001 r9:00000000 r8:00e4bddc r7:000d0d18 r6:00000080 r5:00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.044211]  r4:00000000
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.046899] [<c008b704>] (SyS_futex) from [<c000f880>] (ret_fast_syscall+0x0/0x3c)
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.054557]  r10:00000000 r9:cde52000 r8:c000fa24 r7:000000f0 r6:00000000 r5:000d0d18
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.062642]  r4:b55cb8d8
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.065300] Code: 5f766564 5f746573 61666564 5f746c75 (74687465)
Jan 16 11:45:48 Mighty-B6 kernel[198]: [  879.109379] ---[ end trace da7cb998ff81836b ]---
Jan 16 11:45:49 Mighty-B6 server[207]: [SON_DWL] Waiting For Spotify Stop Prefetch Callback
Jan 16 11:45:50 Mighty-B6 server[207]: [SON_DWL] Waiting For Spotify Stop Prefetch Callback
Jan 16 11:45:51 Mighty-B6 server[207]: [SON_DWL] Waiting For Spotify Stop Prefetch Callback
Jan 16 11:45:52 Mighty-B6 server[207]: [SON_DWL] Waiting For Spotify Stop Prefetch Callback

over 7 years ago

0 Biser Gatchev-XID over 7 years ago

TI__Guru**** 393215 points

Hi,

What is your Linux version? Can you post the DDR portion of the schematics?

0 Aniket Jesu over 7 years ago in reply to Biser Gatchev-XID

Prodigy 200 points

Hello Biser,

Here are the details for Linux version:

4.1.18-gbbe8cfc #1 PREEMPT Tue Dec 5 17:22:08 IST 2017 armv7l GNU/Linux

And DDR part of schematics is attached for reference.

thanks,

aniket

0 Biser Gatchev-XID over 7 years ago in reply to Aniket Jesu

TI__Guru**** 393215 points

Schematic seems OK. About the Linux version, is this the Linux-RT Processor SDK or did you take it from some other source?

0 Aniket Jesu over 7 years ago in reply to Biser Gatchev-XID

Prodigy 200 points

We have got linux sources from this TI link: software-dl.ti.com/.../index_FDS.html

This is linux processor sdk from TI, not the linux-RT.

And we have built this branch : processor-sdk-linux-02.00.02 using yocto build with arago-tmp-external-linaro-toolchain.

0 Pavel Botev over 7 years ago in reply to Aniket Jesu

TI__Guru**** 170625 points

Hi Aniket,

From what I understand, you first put AM335x device in low power mode then put DDR is self-refresh mode. Could you please provide console log with these commands?

Note that in linux 4.1.x you can put AM335x device is low power mode from user space, refer to the below wiki:

processors.wiki.ti.com/.../Linux_Core_Power_Management_User's_Guide_(v4.1)

And when you put AM335x in low power state, DDR is set to self-refresh automatically, no need to do that manually. Refer to the below pointers for more info:

AM335x datasheet, Table 5-11. AM335x Low-Power Modes Power Consumption Summary

www.ti.com/.../sprac74a.pdf

processors.wiki.ti.com/.../Processor_SDK_Linux_Kernel_Performance_Guide

Regards,
Pavel

0 Aniket Jesu over 7 years ago in reply to Pavel Botev

Prodigy 200 points

Hello Pavel,

Low power mode we meant not as Sleep/Standby mode.

For battery optimization, we are trying to run the AM3351 at lowest possible operating mode for Audio playback.
For that we are doing the following steps :

1. Reduce the CPU to 100 Mhz ( Using powersave governor, and adding a 100Mhz in DTS)
2. Reduce CORE clock to 25 Mhz (using devmem2 0x44E00480 w 0x228)
3. Set the DDR3 to DDR3 Auto Self Refresh (using devmem2 0x4C000038 w 0x2A0)
4. Reducing the VDD_CORE step by step from 1.15 Volts to 0.975 Volts.

The Linux Crash occurs only upon enabling the DDR auto-self refresh. So, if we skip step 3 ,
then with rest of the steps, there is no crash.

Can you suggest on this ?

thanks,
aniket jesu

0 Pavel Botev over 7 years ago in reply to Aniket Jesu

TI__Guru**** 170625 points

Aniket,

Aniket Jesu said:
For battery optimization, we are trying to run the AM3351 at lowest possible operating mode for Audio playback.

Are you sure that you are not using frequently DDR in this audio playback usecase? The main purpose of the DDR auto self-refresh is when DDR is not used for long periods.

Aniket Jesu said:
3. Set the DDR3 to DDR3 Auto Self Refresh (using devmem2 0x4C000038 w 0x2A0)
4. Reducing the VDD_CORE step by step from 1.15 Volts to 0.975 Volts.

The max supported voltage for VDD_CORE is 1.144V (OPP100).

Seems you are switching from OPP100 (vdd_core at 1.1V) to OPP50 (0.975), and DDR3 at OPP50 is not supported.

Regards,
Pavel

0 Aniket Jesu over 7 years ago in reply to Pavel Botev

Prodigy 200 points

Pavel,

Are you sure that you are not using frequently DDR in this audio playback usecase?
The main purpose of the DDR auto self-refresh is when DDR is not used for long periods.

YES

-------

The max supported voltage for VDD_CORE is 1.144V (OPP100).

Seems you are switching from OPP100 (vdd_core at 1.1V) to OPP50 (0.975), and DDR3 at OPP50 is not supported.

Actually, we are using a PMIC (tps65217) chip for different voltage rail requirements, so we have VDDS_DDR maintained with 1.35 volts.
can you clarify how VDD_CORE & VDDS_DDR are co-related for OPP100 & OPP50 ?

And also the low power mode technique mentioned in my previous posts, we have been using it in 10k plus units without any issue.
Out of these 10k units we are seeing above mentioned issue only approximately in 100 units.

Irrespective of OPP100 or OPP50, this issue is seen only if we put DDR3 to auto self refresh mode, can you clarify on this ?

And we are using DDR3L, just to be clear.

thanks,

aniket jesu

0 Pavel Botev over 7 years ago in reply to Aniket Jesu

TI__Guru**** 170625 points

Aniket,

Aniket Jesu said:
Actually, we are using a PMIC (tps65217) chip for different voltage rail requirements, so we have VDDS_DDR maintained with 1.35 volts.
can you clarify how VDD_CORE & VDDS_DDR are co-related for OPP100 & OPP50 ?

In AM335x device we have VDD_CORE and VDD_MPU domains only. VDDS_DDR is for the DDR3L IO domain.

This is AM335x requirement documented in datasheet SPRS717J. See sections:

5.4 Operating Performance Points (OPPs)

7.3 OPP50 Support

You will find in these sections that DDR3/DDR3L is not operating in OPP50 mode (voltage below 0.988)

Aniket Jesu said:
And also the low power mode technique mentioned in my previous posts, we have been using it in 10k plus units without any issue.
Out of these 10k units we are seeing above mentioned issue only approximately in 100 units.

From what I understand, these 10K+ boards are absolutely the same, and you have few (around 100) boards that report the issue.

Then this seems to be HW malfunction of these few boards, so you need to double check the HW/PCB design of these boards.

Aniket Jesu said:
Irrespective of OPP100 or OPP50, this issue is seen only if we put DDR3 to auto self refresh mode, can you clarify on this ?

I assume you have made successful SW leveling of DDR3 for all the boards. Please also verify your DDR3 is working correctly, you can use CCS EDMA test, see below wiki:

Regards,
Pavel

0 Paul Chacko over 7 years ago in reply to Pavel Botev

Intellectual 623 points

Hello Pavel,

Thanks for the support. Two questions.

(1) Is there a possibility that the CORE voltage becoming lower can cause some internal damage to the processor's DDR interface?
(2) Are there any other thing to be checked (For example, internal clocks changing during LPM mode)?

Regarding the SW leveling of DDR3, this was done before going to production.

Thanks and regards,

Paul

0 Pavel Botev over 7 years ago in reply to Paul Chacko

TI__Guru**** 170625 points

Paul,

Paul Chacko said:
(1) Is there a possibility that the CORE voltage becoming lower can cause some internal damage to the processor's DDR interface?

When CORE voltage get below 0.988V (OPP50), DDR3L will not operate correct, but I think DDR interface will not be damaged. Voltage values that can damage AM335x device are listed in datasheet, section 5.1 Absolute Maximum Ratings. In this section is stated that CORE voltage below -0.5V can damage AM335x device.

Paul Chacko said:
(2) Are there any other thing to be checked (For example, internal clocks changing during LPM mode)?

Check the below wikis:

Regards,
Pavel

0 Paul Chacko over 7 years ago in reply to Pavel Botev

Intellectual 623 points

Thank you Pavel for those Wiki links.

(1) So the final answer is OPP50 mode is not supported for DDR3L. Correct?

(2) If YES, could you please specify what is the lowest possible CORE Voltage at which DDR3L will operate properly?

(3) How does CORE voltage affects the DDR3L operation?

Thanks and regards,
Paul

0 Pavel Botev over 7 years ago in reply to Paul Chacko

TI__Guru**** 170625 points

Paul,

Paul Chacko said:
(1) So the final answer is OPP50 mode is not supported for DDR3L. Correct?

yes, correct

Paul Chacko said:
(2) If YES, could you please specify what is the lowest possible CORE Voltage at which DDR3L will operate properly?

OPP100 min value, which is 1.056V

Paul Chacko said:
(3) How does CORE voltage affects the DDR3L operation?

The minimum operating frequency of 303MHZ for DDR3 devices requires VDD_CORE to be OPP100. It is not possible to operate the AM335x DDR interface at at 303MHz when sourcing VDD_CORE with the OPP50 defined voltage range.

the JEDEC spec only goes down to 303MHz for DDR3 devices. if you look at DDR3 memory specs they say they do not guarantee operation below 303MHz. we cannot support 303MHz with the core voltage at .95V which is why we don't list OPP50 with DDR3.

Regards,
Pavel

Processors

Processors forum

Linux/AM3351: DDR self-refresh issues