Description of problem: I sucessfully installed FC5/i686 on the MPT SAS drive on an HP xw9400, but needed the "noapic" parameter to boot and install. After the install, a normal SMP kernel boots fine. I installed the latest FC5 xen0 kernel (which is kernel-xen0-2.6.16-1.2122_FC5), and rebooted. During the boot of this kernel, while bringing up the SAS driver, I get the following stack trace: mptbase: Initiating ioc0 bringup ioc0: SAS1068E: Capabilities={Initiator} mptbase: Initiating ioc0 recovery BUG: soft lockup detected on CPU#2! <c043ed9e> softlockup_tick+0x9f/0xb4 <c0408727> timer_interrupt+0x4fc/0x545 <c043ee59> handle_IRQ_event+0x42/0x85 <c043ef29> __do_IRQ+0x8d/0xdc <c04066bc> do_IRQ+0x1a/0x25 <c053d380> evtchn_do_upcall+0x66/0x9f <c0404d91> hypervisor_callback+0x3d/0x48 <c0407219> delay_tsc+0xd/0x15 <f483dadb> WaitForDoorbellInt+0x97/0xb8 [mptbase] <f483dcc1> mpt_handshake_req_reply_wait+0x119/0x25a [mptbase] <f483e3f2> SendIocInit+0x213/0x2c6 [mptbase] <f4842204> mpt_do_ioc_recovery+0xe64/0xe96 [mptbase] <c04d78ab> __next_cpu+0x11/0x22 <c041687a> find_busiest_group+0xfe/0x2a0 <c04175fc> rebalance_tick+0x109/0x269 <c0408760> timer_interrupt+0x535/0x545 <f48422c7> mpt_HardResetHandler+0x91/0xf5 [mptbase] <f4842338> mpt_timer_expired+0xd/0x25 [mptbase] <c0425832> run_timer_softirq+0x122/0x17c <f484232b> mpt_timer_expired+0x0/0x25 [mptbase] <c0421727> __do_softirq+0x70/0xef <c04217e6> do_softirq+0x40/0x67 <c04066c1> do_IRQ+0x1f/0x25 <c053d380> evtchn_do_upcall+0x66/0x9f <c0404d91> hypervisor_callback+0x3d/0x48 <c0402bf5> xen_idle+0x60/0x7f <c0402d53> cpu_idle+0xbc/0xd5 This repeats a few times (anywhere from 1-5 more times), and then it gets by, and boots the rest of the kernel fine. I also tried the latest fc6 kernel (kernel-xen0-2.6.16-1.2236_FC6), with the same result. Version-Release number of selected component (if applicable): How reproducible: Always. Steps to Reproduce: 1. Install FC-5/i686 on an HP xw9400 2. Install the kernel-xen0 kernel. 3. Reboot into kernel-xen0 Actual results: Stack trace as described above. Expected results: Normal boot. Additional info:
change QA contact
This report targets FC5, which is now end-of-life. Please re-test against Fedora 7 or later, and if the issue persists, open a new bug. Thanks