Intel Virtualization and clock_gettime reproducer

CK_REALTIME, &ts); end = GetPerf(); delta = end - start; if (delta > max) max = delta; if (delta < min) min = delta; total += delta; } avg = total / count; printf("min: %ld\n", min); -- //E. F. Berkley Shands, MSc Senior Software

VDSO and Intel Skylake Virtualization

re is a hypervisor before getting confused in __vdso_clock_gettime() ? Berkley -- //E. F. Berkley Shands, MSc Senior Software Architect/Engineer Exegy Systems Engineering// **Exegy Inc.** 349 Marshall Road, Suite 100 St. Louis , MO 63119 Direct: (314) 218-3600 X450 Cell: (314) 303-2546 Offic

NUMA allocations fail to be numa allocated

using libnuma calls on RedHat 6.3 x86_64 with the default kernel and up to 3.4.29 don't allocate on the specified numa nodes, even when forced with numactl. It appears that setting the NUMA policy, and or numa nodes does little for large allocations. Using HUGETLBFS, and you get memory on most

Re: 2.6.23 spinlock hang in kswapd under heavy disk write loads

With DEBUG_SLAB on, I can run only a very short time under 2.6.23 before a kernel panic. [ 626.028180] eth0: too many iterations (6) in nv_nic_irq. [ 626.167583] eth0: too many iterations (6) in nv_nic_irq. [ 626.206729] eth0: too many iterations (6) in nv_nic_irq. [ 626.400171] eth0: too man

Re: 2.6.23 spinlock hang in kswapd under heavy disk write loads

100% reproducible on the two motherboards in question. Does not happen on any other motherboard I have in my possession (not tyan, not uniwide, not socket 940...) No errors, no dmesg, nothing with debug_spinlock set. shows lots (when it works), but by then too many things are locked up to be of

2.6.23 spinlock hang in kswapd under heavy disk write loads

2.6.23 with CONFIG_DEBUG_SPINLOCK on does not hang under very high write loads to either an LSIELP (write rate 1.1GB/Sec) or to a highpoint RR2340 (write rate 1.0GB/Sec). With CONFIG_DEBUG_SPINLOCK off however, the system hangs with kswapd getting 100% of the cpu and most if not all other proc

2.6.23 gotcha megaraid_sas missing

Kconfig in drivers/scsi/Kconfig is wrong. it should default to on. menuconfig SCSI_LOWLEVEL bool "SCSI low-level drivers" depends on SCSI!=n berkley -- // E. F. Berkley Shands, MSc// ** Exegy Inc.** 349 Marshall Road, Suite 100 St. Louis , MO 63119 Direct: (314)

2.6.23-rc9 kswapd infinite loop

nough to get the following ps, top, and /proc/meminfo data Hints anyone (please) as to how to slay this dragon? Berkley -- // E. F. Berkley Shands, MSc// ** Exegy Inc.** 349 Marshall Road, Suite 100 St. Louis , MO 63119 Direct: (314) 218-3600 X450 Cell: (314) 303-2546 Office: (314

2.6.22.5 forcedeth timeout hang

100% reproducible hang on xmit timeout. Just do a "make -j4 modules" on an nfs mounted kernel source. attached is the messages log berkley -- // E. F. Berkley Shands, MSc// ** Exegy Inc.** 349 Marshall Road, Suite 100 St. Louis , MO 63119 Direct: (314) 218-3600 X450 Cell:

Quick fix for LSI8888ELP overruns in 2.6.20 and 2.6.21-rc4

My local kernel guru recognized this problem as a duplicate of the LSI8408E bus lockup problem. The controller gets flooded with work, and just climbs under a rock to hide. In the 8408E case, it froze the PCI-e bus :-) #!/bin/csh # # set the max request queue down from 128. any more than 32 # qui

2.6.20 (+2.6.21-rc4) write hang modifier

s dropping something. berkley -- //E. F. Berkley Shands, MSc// **Exegy Inc.** 3668 S. Geyer Road, Suite 300 St. Louis, MO 63127 Direct: (314) 450-5348 Cell: (314) 303-2546 Office: (314) 450-5353 Fax: (314) 450-5354 The Usual Disclaimer follows... This e-mail and any documents accom

2.6.19 xfs crash spawns tcp reclen errors off nfs

wth should have been fixed in 2.6.18-rc5, but it still happens, easily triggered by the XFS shutdown. the packet sizes reported can grow to > 1GB. Dropping back to 2.6.18.1 stops the XFS crashing, and hence stops the NFS mess. berkley -- //E. F. Berkley Shands, MSc// **Exegy Inc.** 3668

Devices/Partitions over 2TB

With a Broadcom BC4852 and suitable Sata drives, it is easy to create functional devices with well in excess of 2TB raw space. This presents a severe problem to partitioning tools, such as fdisk/cfdisk and the like as the kernel partition structure has a 32 bit integer max for sector counts

2.6.10 Kernel BUG at hugetlbpage:212 (x86_64 and i386)

Reproducible BUG on 3GB hugetlbfs filesystem for opterons and xeons with either FC3 or RedHat ES3.0 and GCC 3.4.2. Details and code snippets in attachment. Executables to reproduce BUG are available on request. berkley On an 8GB dual cpu opteron (Tyan S2884) 2.6.10 kernel I can reproduce a crash

/proc/net/atm/ handling wrong 2.4.0-test9+

This has been reported in the linux-atm mailing list as well.. /usr/src/linux/net/atm/proc.c has a problem. in 2.4.0-test9 and later, the routine atm_proc_dev_register(); is called before atm_proc_init(); the latter is called at the very end of startup, long after the pci devices are initialized.