Bug 19622 - java Process Which Uses HugePages (the UseLargePages jvm option) dies every few hours
Summary: java Process Which Uses HugePages (the UseLargePages jvm option) dies every f...
Status: RESOLVED WILL_NOT_FIX
Alias: None
Product: Memory Management
Classification: Unclassified
Component: Page Allocator (show other bugs)
Hardware: All Linux
: P1 high
Assignee: Andrew Morton
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-10-03 07:49 UTC by yonadav
Modified: 2010-10-04 21:16 UTC (History)
0 users

See Also:
Kernel Version: 2.6.18-128.el5
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description yonadav 2010-10-03 07:49:44 UTC
The java process is attempting to utilize HugePages (via UseLargePages jvm option) but keeps crashing every few hours. The machine is running on VMWare ESXi with VMware tools installed but with the vmmem driver module removed (it was behaving badly due to an issue it has with huge pages as well).

In /var/log/messages you may find:
Sep 30 14:47:06 localhost kernel: BUG: soft lockup - CPU#1 stuck for 10s! [java:27851]
Sep 30 14:47:06 localhost kernel: CPU 1:
Sep 30 14:47:07 localhost kernel: Modules linked in: vsock(U) vmci(U) ipv6 xfrm_nalgo crypto_api dm_mirror dm_multipath scsi_dh video hwmon backlight sbs i2c_ec button battery asus_acpi acpi_memhotplug ac pvscsi(U) lp sg floppy shpchp serio_raw parport_pc i2c_piix4 ide_cd parport i2c_core cdrom e1000 pcspkr dm_raid45 dm_message dm_region_hash dm_log dm_mod dm_mem_cache ata_piix libata mptspi mptscsih mptbase scsi_transport_spi sd_mod scsi_mod ext3 jbd uhci_hcd ohci_hcd ehci_hcd
Sep 30 14:47:07 localhost kernel: Pid: 27851, comm: java Tainted: G      2.6.18-128.el5 #1
Sep 30 14:47:07 localhost kernel: RIP: 0010:[<ffffffff8000c6bc>]  [<ffffffff8000c6bc>] put_page+0x0/0x2e
Sep 30 14:47:07 localhost kernel: RSP: 0018:ffff810013b33bd0  EFLAGS: 00010286
Sep 30 14:47:07 localhost kernel: RAX: 0000000000dc8000 RBX: ffff8100db421740 RCX: ffff810001000038
Sep 30 14:47:07 localhost kernel: RDX: 0000000000000007 RSI: 000000000003f000 RDI: ffff810001509000
Sep 30 14:47:07 localhost kernel: RBP: 0000000000000000 R08: 000000001a02b067 R09: 0000000000036a4f
Sep 30 14:47:07 localhost kernel: R10: 00002aaaaabcc000 R11: 0000000000000000 R12: ffffffff8000aed4
Sep 30 14:47:07 localhost kernel: R13: ffff810003a3e878 R14: ffff81000000db00 R15: ffff810001334070
Sep 30 14:47:07 localhost kernel: FS:  0000000000000000(0000) GS:ffff810037d0f840(0000) knlGS:0000000000000000
Sep 30 14:47:07 localhost kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
Sep 30 14:47:07 localhost kernel: CR2: 00002abffe58d000 CR3: 0000000000201000 CR4: 00000000000006e0
Sep 30 14:47:07 localhost kernel:
Sep 30 14:47:07 localhost kernel: Call Trace:
Sep 30 14:47:07 localhost kernel:  [<ffffffff800d0b1a>] __unmap_hugepage_range+0x10f/0x144
Sep 30 14:47:07 localhost kernel:  [<ffffffff800d0d7d>] unmap_hugepage_range+0x37/0x54
Sep 30 14:47:08 localhost kernel:  [<ffffffff800077dd>] unmap_vmas+0xf8/0x848
Sep 30 14:47:08 localhost kernel:  [<ffffffff80039aad>] exit_mmap+0x78/0xf3
Sep 30 14:47:08 localhost kernel:  [<ffffffff8003bc07>] mmput+0x30/0x83
Sep 30 14:47:08 localhost kernel:  [<ffffffff800152f8>] do_exit+0x2b1/0x91f
Sep 30 14:47:09 localhost kernel:  [<ffffffff80048c18>] cpuset_exit+0x0/0x6c
Sep 30 14:47:09 localhost kernel:  [<ffffffff8002ad05>] get_signal_to_deliver+0x42c/0x45a
Sep 30 14:47:09 localhost kernel:  [<ffffffff8005a837>] do_notify_resume+0x9c/0x7af
Sep 30 14:47:09 localhost kernel:  [<ffffffff8008a461>] default_wake_function+0x0/0xe
Sep 30 14:47:09 localhost kernel:  [<ffffffff8000d0d4>] dput+0x2c/0x114
Sep 30 14:47:09 localhost kernel:  [<ffffffff800a115e>] sys_futex+0x10b/0x12b
Sep 30 14:47:09 localhost kernel:  [<ffffffff8005d19f>] sysret_signal+0x1c/0x27
Sep 30 14:47:10 localhost kernel:  [<ffffffff8005d427>] ptregscall_common+0x67/0xac
Sep 30 14:47:10 localhost kernel:
Comment 1 Andrew Morton 2010-10-04 21:16:17 UTC
2.6.18 is terribly old.  Please take this up with redhat?

Note You need to log in before you can comment on or make changes to this bug.