Skip to contentSkip to conversation

Linux 2.6.35.6

DISCUSSION Greg KH 5 participants 21 messages Sep 27, 2010 – Sep 30, 2010

Subsystems (65)
lore
Share & tools
lore

Greg KH

author
#0 lore raw

I'm announcing the release of the 2.6.35.6 kernel.

All users of the 2.6.35 kernel series must upgrade.

The updated 2.6.35.y git tree can be found at:

        git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.35.y.git
and can be browsed at the normal kernel.org git web browser:
        http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.35.y.git;a=summary

thanks,

greg k-h

------------

 Makefile                                 |    2 
 arch/alpha/kernel/err_marvel.c           |    6 
 arch/alpha/kernel/proto.h                |    3 
 arch/alpha/kernel/sys_cabriolet.c        |   19 +
 arch/alpha/kernel/sys_takara.c           |   11 -
 arch/arm/mach-at91/at91sam9g45_devices.c |    2 
 arch/ia64/kernel/fsys.S                  |   30 ++
 arch/sparc/include/asm/oplib_64.h        |   27 --
 arch/sparc/prom/cif.S                    |   16 -
 arch/sparc/prom/console_64.c             |   48 +++-
 arch/sparc/prom/devops_64.c              |   36 +++
 arch/sparc/prom/misc_64.c                |  314 +++++++++++++++++++++----------
 arch/sparc/prom/p1275.c                  |  102 ----------
 arch/sparc/prom/tree_64.c                |  210 +++++++++++++++-----
 arch/x86/kvm/emulate.c                   |    4 
 arch/x86/kvm/mmu.c                       |    8 
 arch/x86/kvm/paging_tmpl.h               |   28 ++
 arch/x86/kvm/vmx.c                       |    4 
 arch/x86/kvm/x86.c                       |    7 
 arch/x86/lib/atomic64_386_32.S           |   38 +--
 arch/x86/oprofile/nmi_int.c              |    4 
 drivers/base/power/main.c                |    1 
 drivers/char/agp/intel-agp.c             |   44 +++-
 drivers/char/agp/intel-agp.h             |   19 +
 drivers/char/agp/intel-gtt.c             |   47 ++++
 drivers/char/mem.c                       |    3 
 drivers/char/virtio_console.c            |    6 
 drivers/gpu/drm/i915/i915_drv.c          |    6 
 drivers/gpu/drm/i915/intel_display.c     |    8 
 drivers/gpu/drm/nouveau/nv50_instmem.c   |    2 
 drivers/message/fusion/mptscsih.c        |    2 
 drivers/net/bonding/bond_3ad.c           |    3 
 drivers/net/bonding/bond_alb.c           |    3 
 drivers/net/cxgb3/cxgb3_main.c           |    2 
 drivers/net/eql.c                        |    2 
 drivers/net/usb/hso.c                    |    2 
 drivers/pci/intel-iommu.c                |   90 ++++----
 drivers/platform/x86/dell-wmi.c          |    2 
 drivers/platform/x86/hp-wmi.c            |   51 +++--
 drivers/rtc/rtc-s3c.c                    |   13 -
 drivers/staging/vt6655/wpactl.c          |   11 -
 drivers/usb/musb/musb_debugfs.c          |    5 
 drivers/usb/serial/mos7720.c             |    3 
 drivers/usb/serial/mos7840.c             |    3 
 drivers/video/sis/sis_main.c             |    3 
 drivers/video/via/ioctl.c                |    2 
 fs/aio.c                                 |   13 +
 fs/char_dev.c                            |    4 
 fs/compat.c                              |    2 
 fs/gfs2/log.c                            |    2 
 fs/notify/inotify/inotify_user.c         |    7 
 fs/proc/vmcore.c                         |    2 
 fs/xfs/linux-2.6/xfs_ioctl.c             |    2 
 include/linux/ksm.h                      |   20 -
 include/linux/kvm_host.h                 |    1 
 include/linux/mm.h                       |    8 
 include/linux/mmzone.h                   |   13 +
 include/linux/vmstat.h                   |   22 ++
 include/net/sock.h                       |    1 
 include/net/tcp.h                        |   36 ++-
 include/net/udp.h                        |    1 
 kernel/exit.c                            |    5 
 kernel/hw_breakpoint.c                   |    3 
 kernel/power/snapshot.c                  |   85 ++++++--
 kernel/sched.c                           |    8 
 kernel/sys.c                             |    2 
 mm/backing-dev.c                         |    2 
 mm/ksm.c                                 |    3 
 mm/memory.c                              |   56 ++++-
 mm/mmap.c                                |    4 
 mm/mmzone.c                              |   21 ++
 mm/page_alloc.c                          |   33 ++-
 mm/percpu.c                              |    2 
 mm/vmscan.c                              |   43 +++-
 mm/vmstat.c                              |   15 +
 net/Kconfig                              |    2 
 net/bridge/br_netfilter.c                |    6 
 net/core/skbuff.c                        |    6 
 net/ipv4/datagram.c                      |    5 
 net/ipv4/route.c                         |    7 
 net/ipv4/tcp.c                           |   32 ---
 net/ipv4/tcp_timer.c                     |    8 
 net/ipv4/udp.c                           |   44 ++++
 net/ipv6/datagram.c                      |    7 
 net/ipv6/udp.c                           |   10 
 net/irda/af_irda.c                       |    4 
 net/l2tp/l2tp_eth.c                      |    2 
 net/llc/af_llc.c                         |    3 
 net/rds/recv.c                           |    2 
 net/sctp/output.c                        |    1 
 net/unix/af_unix.c                       |   15 +
 security/keys/keyctl.c                   |    6 
 sound/pci/hda/patch_sigmatel.c           |   12 +
 virt/kvm/kvm_main.c                      |    1 
 94 files changed, 1264 insertions(+), 577 deletions(-)

Alexey Kuznetsov (1):
tcp: Prevent overzealous packetization by SWS logic.

Amit Shah (1):
virtio: console: Prevent userspace from submitting NULL buffers

Andrea Arcangeli (3):
mm: fix swapin race condition
mmap: call unlink_anon_vmas() in __split_vma() in case of error
KVM: MMU: fix mmu notifier invalidate handler for huge spte

Andrew Morton (1):
drivers/pci/intel-iommu.c: fix build with older gcc's

Andy Gospodarek (1):
bonding: correctly process non-linear skbs

Arnd Bergmann (1):
/proc/vmcore: fix seeking

Avi Kivity (3):
KVM: Keep slot ID in memory slot structure
KVM: Prevent internal slots from being COWed
KVM: VMX: Fix host GDT.LIMIT corruption

Chris Wilson (2):
drm/i915,agp/intel: Add second set of PCI-IDs for B43 drm/i915: Ensure that the crtcinfo is populated during mode_fixup()

Christoph Lameter (1):
mm: page allocator: calculate a better estimate of NR_FREE_PAGES when memory is low and kswapd is awake

Colin Cross (1):
PM: Prevent waiting forever on asynchronous resume after failing suspend

Dan Carpenter (2):
Staging: vt6655: fix buffer overflow net/llc: make opt unsigned in llc_ui_setsockopt()

Dan Rosenberg (8):
USB: serial/mos*: prevent reading uninitialized stack memory drivers/net/usb/hso.c: prevent reading uninitialized memory drivers/net/cxgb3/cxgb3_main.c: prevent reading uninitialized stack memory drivers/net/eql.c: prevent reading uninitialized stack memory drivers/video/sis/sis_main.c: prevent reading uninitialized stack memory Prevent freeing uninitialized pointer in compat_do_readv_writev
xfs: prevent reading uninitialized stack memory drivers/video/via/ioctl.c: prevent reading uninitialized stack memory

Daniel J Blueman (2):
ALSA: hda - Fix beep frequency on IDT 92HD73xx and 92HD71Bxx codecs Fix unprotected access to task credentials in waitid()

David Howells (2):
KEYS: Fix RCU no-lock warning in keyctl_session_to_parent()
KEYS: Fix bug in keyctl_session_to_parent() if parent has no session keyring

David S. Miller (5):
bridge: Clear INET control block of SKBs passed into ip_fragment().
irda: Correctly clean up self->ias_obj on irda_bind() failure.
net: RPS needs to depend upon USE_GENERIC_SMP_HELPERS
tcp: Combat per-cpu skew in orphan tests.
sparc64: Get rid of indirect p1275 PROM call buffer.

Eric Dumazet (5):
gro: fix different skb headrooms
rds: fix a leak of kernel memory
tcp: fix three tcp sysctls tuning
udp: add rehash on connect()
l2tp: test for ethernet header in l2tp_eth_dev_recv()

Eric Paris (1):
inotify: send IN_UNMOUNT events

Gleb Natapov (1):
KVM: x86: emulator: inc/dec can have lock prefix

Greg Kroah-Hartman (1):

      Linux 2.6.35.6

Hans de Goede (1):
virtio: console: Fix poll blocking even though there is data to read

Hugh Dickins (1):
mm: further fix swapin race condition

Islam Amer (1):
dell-wmi: Add support for eject key on Dell Studio 1555

Jan Kara (3):
bdi: Initialize noop_backing_dev_info properly
char: Mark /dev/zero and /dev/kmem as not capable of writeback
aio: do not return ERESTARTSYS as a result of AIO

Jarek Poplawski (1):
gro: Re-fix different skb headrooms

Jeff Moyer (1):
aio: check for multiplication overflow in do_io_submit

Jianzhao Wang (1):
net: blackhole route should always be recalculated

KOSAKI Motohiro (1):
tcp: select(writefds) don't hang up when a peer close connection

Luca Barbieri (1):
x86, asm: Use a lower case name for the end macro in atomic64_386_32.S

Marcin Slusarz (1):
drm/nv50: initialize ramht_refs list for faked 0 channel

Mathias Nyman (1):
usb: musb_debugfs: don't use the struct file private_data field with seq_files

Matt Helsley (1):
hw breakpoints: Fix pid namespace bug

Mel Gorman (2):
mm: page allocator: drain per-cpu lists after direct reclaim allocation fails
mm: page allocator: update free page counters after pages are placed on the free list

Michael Cree (1):
alpha: Fix printk format errors

Minchan Kim (1):
vmscan: check all_unreclaimable in direct reclaim path

Morten H. Larsen (1):
Fix call to replaced SuperIO functions

Nicolas Ferre (1):
AT91: change dma resource index

Patrick Simmons (1):
oprofile: Add Support for Intel CPU Family 6 / Model 22 (Intel Celeron 540)

Paul E. McKenney (1):
pid: make setpgid() system call use RCU read-side critical section

Petr Tesarik (1):
IA64: Optimize ticket spinlocks in fsys_rt_sigprocmask

Rafael J. Wysocki (1):
PM / Hibernate: Avoid hitting OOM during preallocation of memory

Ryan Kuester (1):
SCSI: mptsas: fix hangs caused by ATA pass-through

Stanislaw Gruszka (1):
sched: Fix user time incorrectly accounted as system time on 32-bit

Steven Whitehouse (1):
GFS2: gfs2_logd should be using interruptible waits

Tejun Heo (1):
percpu: fix pcpu_last_unit_cpu

Tetsuo Handa (1):
UNIX: Do not loop forever at unix_autobind().

Thomas Renninger (1):
x86 platform drivers: hp-wmi Reorder event id processing

Tony Luck (2):
IA64: fix siglock
guard page for stacks that grow upwards

Vlad Yasevich (1):
sctp: Do not reset the packet during sctp_packet_config().

Vladimir Zapolskiy (1):
rtc: s3c: balance state changes of wakeup flag

Xiao Guangrong (1):
KVM: MMU: fix direct sp's access corrupted

Zhenyu Wang (3):
intel_agp, drm/i915: Add all sandybridge graphics devices support agp/intel: fix physical address mask bits for sandybridge agp/intel: fix dma mask bits on sandybridge

author
#1 lore raw
Makefile +1 -1
index 11aaa51..78f9271 100644
@@ -1,7 +1,7 @@
 VERSION = 2
 PATCHLEVEL = 6
 SUBLEVEL = 35
-EXTRAVERSION = .5
+EXTRAVERSION = .6
 NAME = Sheep on Meth
 
 # *DOCUMENTATION*
arch/alpha/kernel/err_marvel.c +3 -3
index 52a79df..5c905aa 100644
@@ -109,7 +109,7 @@ marvel_print_err_cyc(u64 err_cyc)
 #define IO7__ERR_CYC__CYCLE__M	(0x7)
 
 	printk("%s        Packet In Error: %s\n"
-	       "%s        Error in %s, cycle %ld%s%s\n",
+	       "%s        Error in %s, cycle %lld%s%s\n",
 	       err_print_prefix, 
 	       packet_desc[EXTRACT(err_cyc, IO7__ERR_CYC__PACKET)],
 	       err_print_prefix,
@@ -313,7 +313,7 @@ marvel_print_po7_ugbge_sym(u64 ugbge_sym)
 	}
 
 	printk("%s      Up Hose Garbage Symptom:\n"
-	       "%s        Source Port: %ld - Dest PID: %ld - OpCode: %s\n", 
+	       "%s        Source Port: %lld - Dest PID: %lld - OpCode: %s\n",
 	       err_print_prefix,
 	       err_print_prefix, 
 	       EXTRACT(ugbge_sym, IO7__PO7_UGBGE_SYM__UPH_SRC_PORT),
@@ -552,7 +552,7 @@ marvel_print_pox_spl_cmplt(u64 spl_cmplt)
 #define IO7__POX_SPLCMPLT__REM_BYTE_COUNT__M	(0xfff)
 
 	printk("%s      Split Completion Error:\n"	
-	       "%s         Source (Bus:Dev:Func): %ld:%ld:%ld\n",
+	       "%s         Source (Bus:Dev:Func): %lld:%lld:%lld\n",
 	       err_print_prefix,
 	       err_print_prefix,
 	       EXTRACT(spl_cmplt, IO7__POX_SPLCMPLT__SOURCE_BUS),
arch/alpha/kernel/proto.h +0 -3
index 3d2627e..d3e52d3 100644
@@ -156,9 +156,6 @@ extern void SMC669_Init(int);
 /* es1888.c */
 extern void es1888_init(void);
 
-/* ns87312.c */
-extern void ns87312_enable_ide(long ide_base);
-
 /* ../lib/fpreg.c */
 extern void alpha_write_fp_reg (unsigned long reg, unsigned long val);
 extern unsigned long alpha_read_fp_reg (unsigned long reg);
arch/alpha/kernel/sys_cabriolet.c +16 -3
index affd0f3..14c8898 100644
@@ -33,7 +33,7 @@
 #include "irq_impl.h"
 #include "pci_impl.h"
 #include "machvec_impl.h"
-
+#include "pc873xx.h"
 
 /* Note mask bit is true for DISABLED irqs.  */
 static unsigned long cached_irq_mask = ~0UL;
@@ -236,17 +236,30 @@ cabriolet_map_irq(struct pci_dev *dev, u8 slot, u8 pin)
 }
 
 static inline void __init
+cabriolet_enable_ide(void)
+{
+	if (pc873xx_probe() == -1) {
+		printk(KERN_ERR "Probing for PC873xx Super IO chip failed.\n");
+	 } else {
+		printk(KERN_INFO "Found %s Super IO chip at 0x%x\n",
+			pc873xx_get_model(), pc873xx_get_base());
+
+		pc873xx_enable_ide();
+	}
+}
+
+static inline void __init
 cabriolet_init_pci(void)
 {
 	common_init_pci();
-	ns87312_enable_ide(0x398);
+	cabriolet_enable_ide();
 }
 
 static inline void __init
 cia_cab_init_pci(void)
 {
 	cia_init_pci();
-	ns87312_enable_ide(0x398);
+	cabriolet_enable_ide();
 }
 
 /*
arch/alpha/kernel/sys_takara.c +9 -2
index 2304648..4da596b 100644
@@ -29,7 +29,7 @@
 #include "irq_impl.h"
 #include "pci_impl.h"
 #include "machvec_impl.h"
-
+#include "pc873xx.h"
 
 /* Note mask bit is true for DISABLED irqs.  */
 static unsigned long cached_irq_mask[2] = { -1, -1 };
@@ -264,7 +264,14 @@ takara_init_pci(void)
 		alpha_mv.pci_map_irq = takara_map_irq_srm;
 
 	cia_init_pci();
-	ns87312_enable_ide(0x26e);
+
+	if (pc873xx_probe() == -1) {
+		printk(KERN_ERR "Probing for PC873xx Super IO chip failed.\n");
+	} else {
+		printk(KERN_INFO "Found %s Super IO chip at 0x%x\n",
+			pc873xx_get_model(), pc873xx_get_base());
+		pc873xx_enable_ide();
+	}
 }
 
 
arch/arm/mach-at91/at91sam9g45_devices.c +1 -1
index 809114d..501ac6f 100644
@@ -46,7 +46,7 @@ static struct resource hdmac_resources[] = {
 		.end	= AT91_BASE_SYS + AT91_DMA + SZ_512 - 1,
 		.flags	= IORESOURCE_MEM,
 	},
-	[2] = {
+	[1] = {
 		.start	= AT91SAM9G45_ID_DMA,
 		.end	= AT91SAM9G45_ID_DMA,
 		.flags	= IORESOURCE_IRQ,
arch/ia64/kernel/fsys.S +21 -9
index 3567d54..331d42b 100644
@@ -420,22 +420,31 @@ EX(.fail_efault, ld8 r14=[r33])			// r14 <- *set
 	;;
 
 	RSM_PSR_I(p0, r18, r19)			// mask interrupt delivery
-	mov ar.ccv=0
 	andcm r14=r14,r17			// filter out SIGKILL & SIGSTOP
+	mov r8=EINVAL			// default to EINVAL
 
 #ifdef CONFIG_SMP
-	mov r17=1
+	// __ticket_spin_trylock(r31)
+	ld4 r17=[r31]
 	;;
-	cmpxchg4.acq r18=[r31],r17,ar.ccv	// try to acquire the lock
-	mov r8=EINVAL			// default to EINVAL
+	mov.m ar.ccv=r17
+	extr.u r9=r17,17,15
+	adds r19=1,r17
+	extr.u r18=r17,0,15
+	;;
+	cmp.eq p6,p7=r9,r18
 	;;
+(p6)	cmpxchg4.acq r9=[r31],r19,ar.ccv
+(p6)	dep.z r20=r19,1,15		// next serving ticket for unlock
+(p7)	br.cond.spnt.many .lock_contention
+	;;
+	cmp4.eq p0,p7=r9,r17
+	adds r31=2,r31
+(p7)	br.cond.spnt.many .lock_contention
 	ld8 r3=[r2]			// re-read current->blocked now that we hold the lock
-	cmp4.ne p6,p0=r18,r0
-(p6)	br.cond.spnt.many .lock_contention
 	;;
 #else
 	ld8 r3=[r2]			// re-read current->blocked now that we hold the lock
-	mov r8=EINVAL			// default to EINVAL
 #endif
 	add r18=IA64_TASK_PENDING_OFFSET+IA64_SIGPENDING_SIGNAL_OFFSET,r16
 	add r19=IA64_TASK_SIGNAL_OFFSET,r16
@@ -490,7 +499,9 @@ EX(.fail_efault, ld8 r14=[r33])			// r14 <- *set
 (p6)	br.cond.spnt.few 1b			// yes -> retry
 
 #ifdef CONFIG_SMP
-	st4.rel [r31]=r0			// release the lock
+	// __ticket_spin_unlock(r31)
+	st2.rel [r31]=r20
+	mov r20=0					// i must not leak kernel bits...
 #endif
 	SSM_PSR_I(p0, p9, r31)
 	;;
@@ -512,7 +523,8 @@ EX(.fail_efault, (p15) st8 [r34]=r3)
 
 .sig_pending:
 #ifdef CONFIG_SMP
-	st4.rel [r31]=r0			// release the lock
+	// __ticket_spin_unlock(r31)
+	st2.rel [r31]=r20			// release the lock
 #endif
 	SSM_PSR_I(p0, p9, r17)
 	;;
arch/sparc/include/asm/oplib_64.h +3 -24
index a5db031..3e0b2d6 100644
@@ -185,9 +185,8 @@ extern int prom_getunumber(int syndrome_code,
 			   char *buf, int buflen);
 
 /* Retain physical memory to the caller across soft resets. */
-extern unsigned long prom_retain(const char *name,
-				 unsigned long pa_low, unsigned long pa_high,
-				 long size, long align);
+extern int prom_retain(const char *name, unsigned long size,
+		       unsigned long align, unsigned long *paddr);
 
 /* Load explicit I/D TLB entries into the calling processor. */
 extern long prom_itlb_load(unsigned long index,
@@ -287,26 +286,6 @@ extern void prom_sun4v_guest_soft_state(void);
 extern int prom_ihandle2path(int handle, char *buffer, int bufsize);
 
 /* Client interface level routines. */
-extern long p1275_cmd(const char *, long, ...);
-
-#if 0
-#define P1275_SIZE(x) ((((long)((x) / 32)) << 32) | (x))
-#else
-#define P1275_SIZE(x) x
-#endif
-
-/* We support at most 16 input and 1 output argument */
-#define P1275_ARG_NUMBER		0
-#define P1275_ARG_IN_STRING		1
-#define P1275_ARG_OUT_BUF		2
-#define P1275_ARG_OUT_32B		3
-#define P1275_ARG_IN_FUNCTION		4
-#define P1275_ARG_IN_BUF		5
-#define P1275_ARG_IN_64B		6
-
-#define P1275_IN(x) ((x) & 0xf)
-#define P1275_OUT(x) (((x) << 4) & 0xf0)
-#define P1275_INOUT(i,o) (P1275_IN(i)|P1275_OUT(o))
-#define P1275_ARG(n,x) ((x) << ((n)*3 + 8))
+extern void p1275_cmd_direct(unsigned long *);
 
 #endif /* !(__SPARC64_OPLIB_H) */
arch/sparc/prom/cif.S +8 -8
index 5f27ad7..9c86b4b 100644
@@ -9,18 +9,18 @@
 #include <asm/thread_info.h>
 
 	.text
-	.globl	prom_cif_interface
-prom_cif_interface:
-	sethi	%hi(p1275buf), %o0
-	or	%o0, %lo(p1275buf), %o0
-	ldx	[%o0 + 0x010], %o1	! prom_cif_stack
-	save	%o1, -192, %sp
-	ldx	[%i0 + 0x008], %l2	! prom_cif_handler
+	.globl	prom_cif_direct
+prom_cif_direct:
+	sethi	%hi(p1275buf), %o1
+	or	%o1, %lo(p1275buf), %o1
+	ldx	[%o1 + 0x0010], %o2	! prom_cif_stack
+	save	%o2, -192, %sp
+	ldx	[%i1 + 0x0008], %l2	! prom_cif_handler
 	mov	%g4, %l0
 	mov	%g5, %l1
 	mov	%g6, %l3
 	call	%l2
-	 add	%i0, 0x018, %o0		! prom_args
+	 mov	%i0, %o0		! prom_args
 	mov	%l0, %g4
 	mov	%l1, %g5
 	mov	%l3, %g6
arch/sparc/prom/console_64.c +37 -11

70 lines not on this page · raw message

arch/sparc/prom/devops_64.c +31 -5
index 9dbd803..a017119 100644
@@ -18,16 +18,32 @@
 int
 prom_devopen(const char *dstr)
 {
-	return p1275_cmd ("open", P1275_ARG(0,P1275_ARG_IN_STRING)|
-				  P1275_INOUT(1,1),
-				  dstr);
+	unsigned long args[5];
+
+	args[0] = (unsigned long) "open";
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned long) dstr;
+	args[4] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[4];
 }
 
 /* Close the device described by device handle 'dhandle'. */
 int
 prom_devclose(int dhandle)
 {
-	p1275_cmd ("close", P1275_INOUT(1,0), dhandle);
+	unsigned long args[4];
+
+	args[0] = (unsigned long) "close";
+	args[1] = 1;
+	args[2] = 0;
+	args[3] = (unsigned int) dhandle;
+
+	p1275_cmd_direct(args);
+
 	return 0;
 }
 
@@ -37,5 +53,15 @@ prom_devclose(int dhandle)
 void
 prom_seek(int dhandle, unsigned int seekhi, unsigned int seeklo)
 {
-	p1275_cmd ("seek", P1275_INOUT(3,1), dhandle, seekhi, seeklo);
+	unsigned long args[7];
+
+	args[0] = (unsigned long) "seek";
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) dhandle;
+	args[4] = seekhi;
+	args[5] = seeklo;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
 }
arch/sparc/prom/misc_64.c +220 -94

464 lines not on this page · raw message

arch/sparc/prom/p1275.c +3 -99

130 lines not on this page · raw message

arch/sparc/prom/tree_64.c +154 -56

354 lines not on this page · raw message

arch/x86/kvm/emulate.c +2 -2
index 5ac0bb4..715a7fa 100644
@@ -345,10 +345,10 @@ static u32 group_table[] = {
 	DstMem | SrcNone | ModRM, DstMem | SrcNone | ModRM,
 	0, 0, 0, 0,
 	[Group4*8] =
-	ByteOp | DstMem | SrcNone | ModRM, ByteOp | DstMem | SrcNone | ModRM,
+	ByteOp | DstMem | SrcNone | ModRM | Lock, ByteOp | DstMem | SrcNone | ModRM | Lock,
 	0, 0, 0, 0, 0, 0,
 	[Group5*8] =
-	DstMem | SrcNone | ModRM, DstMem | SrcNone | ModRM,
+	DstMem | SrcNone | ModRM | Lock, DstMem | SrcNone | ModRM | Lock,
 	SrcMem | ModRM | Stack, 0,
 	SrcMem | ModRM | Stack, SrcMem | ModRM | Src2Mem16 | ImplicitOps,
 	SrcMem | ModRM | Stack, 0,
arch/x86/kvm/mmu.c +6 -2
index b1ed0a1..92b6ca4 100644
@@ -799,8 +799,12 @@ static int kvm_handle_hva(struct kvm *kvm, unsigned long hva,
 			ret = handler(kvm, &memslot->rmap[gfn_offset], data);
 
 			for (j = 0; j < KVM_NR_PAGE_SIZES - 1; ++j) {
-				int idx = gfn_offset;
-				idx /= KVM_PAGES_PER_HPAGE(PT_DIRECTORY_LEVEL + j);
+				unsigned long idx;
+				int nr;
+
+				nr = KVM_PAGES_PER_HPAGE(PT_DIRECTORY_LEVEL+j);
+				idx = (memslot->base_gfn+gfn_offset) / nr -
+					memslot->base_gfn / nr;
 				ret |= handler(kvm,
 					&memslot->lpage_info[j][idx].rmap_pde,
 					data);
arch/x86/kvm/paging_tmpl.h +26 -2
index 2331bdc..e34452b 100644
@@ -324,8 +324,32 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
 			break;
 		}
 
-		if (is_shadow_present_pte(*sptep) && !is_large_pte(*sptep))
-			continue;
+		if (is_shadow_present_pte(*sptep) && !is_large_pte(*sptep)) {
+			struct kvm_mmu_page *child;
+			unsigned direct_access;
+
+			if (level != gw->level)
+				continue;
+
+			/*
+			 * For the direct sp, if the guest pte's dirty bit
+			 * changed form clean to dirty, it will corrupt the
+			 * sp's access: allow writable in the read-only sp,
+			 * so we should update the spte at this point to get
+			 * a new sp with the correct access.
+			 */
+			direct_access = gw->pt_access & gw->pte_access;
+			if (!is_dirty_gpte(gw->ptes[gw->level - 1]))
+				direct_access &= ~ACC_WRITE_MASK;
+
+			child = page_header(*sptep & PT64_BASE_ADDR_MASK);
+			if (child->role.access == direct_access)
+				continue;
+
+			mmu_page_remove_parent_pte(child, sptep);
+			__set_spte(sptep, shadow_trap_nonpresent_pte);
+			kvm_flush_remote_tlbs(vcpu->kvm);
+		}
 
 		if (is_large_pte(*sptep)) {
 			rmap_remove(vcpu->kvm, sptep);
arch/x86/kvm/vmx.c +4 -0
index ee03679..7d04388 100644
@@ -177,6 +177,7 @@ static u64 construct_eptp(unsigned long root_hpa);
 static DEFINE_PER_CPU(struct vmcs *, vmxarea);
 static DEFINE_PER_CPU(struct vmcs *, current_vmcs);
 static DEFINE_PER_CPU(struct list_head, vcpus_on_cpu);
+static DEFINE_PER_CPU(struct desc_ptr, host_gdt);
 
 static unsigned long *vmx_io_bitmap_a;
 static unsigned long *vmx_io_bitmap_b;
@@ -812,6 +813,7 @@ static void __vmx_load_host_state(struct vcpu_vmx *vmx)
 		wrmsrl(MSR_KERNEL_GS_BASE, vmx->msr_host_kernel_gs_base);
 	}
 #endif
+	load_gdt(&__get_cpu_var(host_gdt));
 }
 
 static void vmx_load_host_state(struct vcpu_vmx *vmx)
@@ -1314,6 +1316,8 @@ static int hardware_enable(void *garbage)
 
 	ept_sync_global();
 
+	store_gdt(&__get_cpu_var(host_gdt));
+
 	return 0;
 }
 
arch/x86/kvm/x86.c +6 -1
index 7fa89c3..fd22688 100644
@@ -5438,6 +5438,11 @@ int kvm_arch_prepare_memory_region(struct kvm *kvm,
 				int user_alloc)
 {
 	int npages = memslot->npages;
+	int map_flags = MAP_PRIVATE | MAP_ANONYMOUS;
+
+	/* Prevent internal slot pages from being moved by fork()/COW. */
+	if (memslot->id >= KVM_MEMORY_SLOTS)
+		map_flags = MAP_SHARED | MAP_ANONYMOUS;
 
 	/*To keep backward compatibility with older userspace,
 	 *x86 needs to hanlde !user_alloc case.
@@ -5450,7 +5455,7 @@ int kvm_arch_prepare_memory_region(struct kvm *kvm,
 			userspace_addr = do_mmap(NULL, 0,
 						 npages * PAGE_SIZE,
 						 PROT_READ | PROT_WRITE,
-						 MAP_PRIVATE | MAP_ANONYMOUS,
+						 map_flags,
 						 0);
 			up_write(&current->mm->mmap_sem);
 
arch/x86/lib/atomic64_386_32.S +20 -18
index 78ee8e0..2cda60a 100644
@@ -26,35 +26,37 @@
 .endm
 
 #define BEGIN(op) \
-.macro END; \
+.macro endp; \
 	CFI_ENDPROC; \
 ENDPROC(atomic64_##op##_386); \
-.purgem END; \
+.purgem endp; \
 .endm; \
 ENTRY(atomic64_##op##_386); \
 	CFI_STARTPROC; \
 	LOCK v;
 
+#define ENDP endp
+
 #define RET \
 	UNLOCK v; \
 	ret
 
-#define RET_END \
+#define RET_ENDP \
 	RET; \
-	END
+	ENDP
 
 #define v %ecx
 BEGIN(read)
 	movl  (v), %eax
 	movl 4(v), %edx
-RET_END
+RET_ENDP
 #undef v
 
 #define v %esi
 BEGIN(set)
 	movl %ebx,  (v)
 	movl %ecx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v  %esi
@@ -63,14 +65,14 @@ BEGIN(xchg)
 	movl 4(v), %edx
 	movl %ebx,  (v)
 	movl %ecx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %ecx
 BEGIN(add)
 	addl %eax,  (v)
 	adcl %edx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %ecx
@@ -79,14 +81,14 @@ BEGIN(add_return)
 	adcl 4(v), %edx
 	movl %eax,  (v)
 	movl %edx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %ecx
 BEGIN(sub)
 	subl %eax,  (v)
 	sbbl %edx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %ecx
@@ -98,14 +100,14 @@ BEGIN(sub_return)
 	adcl 4(v), %edx
 	movl %eax,  (v)
 	movl %edx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %esi
 BEGIN(inc)
 	addl $1,  (v)
 	adcl $0, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %esi
@@ -116,14 +118,14 @@ BEGIN(inc_return)
 	adcl $0, %edx
 	movl %eax,  (v)
 	movl %edx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %esi
 BEGIN(dec)
 	subl $1,  (v)
 	sbbl $0, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %esi
@@ -134,7 +136,7 @@ BEGIN(dec_return)
 	sbbl $0, %edx
 	movl %eax,  (v)
 	movl %edx, 4(v)
-RET_END
+RET_ENDP
 #undef v
 
 #define v %ecx
@@ -156,7 +158,7 @@ BEGIN(add_unless)
 	jne 1b
 	xorl %eax, %eax
 	jmp 2b
-END
+ENDP
 #undef v
 
 #define v %esi
@@ -177,7 +179,7 @@ BEGIN(inc_not_zero)
 	testl %edx, %edx
 	jne 1b
 	jmp 2b
-END
+ENDP
 #undef v
 
 #define v %esi
@@ -190,5 +192,5 @@ BEGIN(dec_if_positive)
 	movl %eax,  (v)
 	movl %edx, 4(v)
 1:
-RET_END
+RET_ENDP
 #undef v
arch/x86/oprofile/nmi_int.c +3 -1
index cfe4faa..009b819 100644
@@ -671,7 +671,9 @@ static int __init ppro_init(char **cpu_type)
 	case 14:
 		*cpu_type = "i386/core";
 		break;
-	case 15: case 23:
+	case 0x0f:
+	case 0x16:
+	case 0x17:
 		*cpu_type = "i386/core_2";
 		break;
 	case 0x1a:
drivers/base/power/main.c +1 -0
index 941fcb8..cb1cedc 100644
@@ -59,6 +59,7 @@ void device_pm_init(struct device *dev)
 {
 	dev->power.status = DPM_ON;
 	init_completion(&dev->power.completion);
+	complete_all(&dev->power.completion);
 	pm_runtime_init(dev);
 }
 
drivers/char/agp/intel-agp.c +31 -13

74 lines not on this page · raw message

drivers/char/agp/intel-agp.h +14 -5
index 2547465..bc9a4ad 100644
@@ -178,6 +178,8 @@
 #define PCI_DEVICE_ID_INTEL_Q33_IG          0x29D2
 #define PCI_DEVICE_ID_INTEL_B43_HB          0x2E40
 #define PCI_DEVICE_ID_INTEL_B43_IG          0x2E42
+#define PCI_DEVICE_ID_INTEL_B43_1_HB        0x2E90
+#define PCI_DEVICE_ID_INTEL_B43_1_IG        0x2E92
 #define PCI_DEVICE_ID_INTEL_GM45_HB         0x2A40
 #define PCI_DEVICE_ID_INTEL_GM45_IG         0x2A42
 #define PCI_DEVICE_ID_INTEL_EAGLELAKE_HB        0x2E00
@@ -194,10 +196,16 @@
 #define PCI_DEVICE_ID_INTEL_IRONLAKE_MA_HB	    0x0062
 #define PCI_DEVICE_ID_INTEL_IRONLAKE_MC2_HB    0x006a
 #define PCI_DEVICE_ID_INTEL_IRONLAKE_M_IG	    0x0046
-#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_HB  0x0100
-#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_IG  0x0102
-#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_HB  0x0104
-#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_IG  0x0106
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_HB		0x0100  /* Desktop */
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_GT1_IG		0x0102
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_GT2_IG		0x0112
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_GT2_PLUS_IG	0x0122
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_HB		0x0104  /* Mobile */
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_GT1_IG	0x0106
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_GT2_IG	0x0116
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_GT2_PLUS_IG	0x0126
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_S_HB		0x0108  /* Server */
+#define PCI_DEVICE_ID_INTEL_SANDYBRIDGE_S_IG		0x010A
 
 /* cover 915 and 945 variants */
 #define IS_I915 (agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_E7221_HB || \
@@ -224,7 +232,8 @@
 		agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_PINEVIEW_HB)
 
 #define IS_SNB (agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_SANDYBRIDGE_HB || \
-		agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_HB)
+		agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_SANDYBRIDGE_M_HB || \
+		agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_SANDYBRIDGE_S_HB)
 
 #define IS_G4X (agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_EAGLELAKE_HB || \
 		agp_bridge->dev->device == PCI_DEVICE_ID_INTEL_Q45_HB || \
drivers/char/agp/intel-gtt.c +45 -2

71 lines not on this page · raw message

drivers/char/mem.c +2 -1
index a398ecd..1f528fa 100644
@@ -788,10 +788,11 @@ static const struct file_operations zero_fops = {
 /*
  * capabilities for /dev/zero
  * - permits private mappings, "copies" are taken of the source of zeros
+ * - no writeback happens
  */
 static struct backing_dev_info zero_bdi = {
 	.name		= "char/mem",
-	.capabilities	= BDI_CAP_MAP_COPY,
+	.capabilities	= BDI_CAP_MAP_COPY | BDI_CAP_NO_ACCT_AND_WRITEBACK,
 };
 
 static const struct file_operations full_fops = {
drivers/char/virtio_console.c +5 -1
index 942a982..c810481 100644
@@ -596,6 +596,10 @@ static ssize_t port_fops_write(struct file *filp, const char __user *ubuf,
 	ssize_t ret;
 	bool nonblock;
 
+	/* Userspace could be out to fool us */
+	if (!count)
+		return 0;
+
 	port = filp->private_data;
 
 	nonblock = filp->f_flags & O_NONBLOCK;
@@ -642,7 +646,7 @@ static unsigned int port_fops_poll(struct file *filp, poll_table *wait)
 	poll_wait(filp, &port->waitqueue, wait);
 
 	ret = 0;
-	if (port->inbuf)
+	if (!will_read_block(port))
 		ret |= POLLIN | POLLRDNORM;
 	if (!will_write_block(port))
 		ret |= POLLOUT;
drivers/gpu/drm/i915/i915_drv.c +6 -0
index 423dc90..194e0c4 100644
@@ -175,12 +175,18 @@ static const struct pci_device_id pciidlist[] = {
 	INTEL_VGA_DEVICE(0x2e22, &intel_g45_info),
 	INTEL_VGA_DEVICE(0x2e32, &intel_g45_info),
 	INTEL_VGA_DEVICE(0x2e42, &intel_g45_info),
+	INTEL_VGA_DEVICE(0x2e92, &intel_g45_info),		/* B43_G.1 */
 	INTEL_VGA_DEVICE(0xa001, &intel_pineview_info),
 	INTEL_VGA_DEVICE(0xa011, &intel_pineview_info),
 	INTEL_VGA_DEVICE(0x0042, &intel_ironlake_d_info),
 	INTEL_VGA_DEVICE(0x0046, &intel_ironlake_m_info),
 	INTEL_VGA_DEVICE(0x0102, &intel_sandybridge_d_info),
+	INTEL_VGA_DEVICE(0x0112, &intel_sandybridge_d_info),
+	INTEL_VGA_DEVICE(0x0122, &intel_sandybridge_d_info),
 	INTEL_VGA_DEVICE(0x0106, &intel_sandybridge_m_info),
+	INTEL_VGA_DEVICE(0x0116, &intel_sandybridge_m_info),
+	INTEL_VGA_DEVICE(0x0126, &intel_sandybridge_m_info),
+	INTEL_VGA_DEVICE(0x010A, &intel_sandybridge_d_info),
 	{0, 0, 0}
 };
 
drivers/gpu/drm/i915/intel_display.c +8 -0
index ac38e46..d1c1243 100644
@@ -2376,11 +2376,19 @@ static bool intel_crtc_mode_fixup(struct drm_crtc *crtc,
 				  struct drm_display_mode *adjusted_mode)
 {
 	struct drm_device *dev = crtc->dev;
+
 	if (HAS_PCH_SPLIT(dev)) {
 		/* FDI link clock is fixed at 2.7G */
 		if (mode->clock * 3 > 27000 * 4)
 			return MODE_CLOCK_HIGH;
 	}
+
+	/* XXX some encoders set the crtcinfo, others don't.
+	 * Obviously we need some form of conflict resolution here...
+	 */
+	if (adjusted_mode->crtc_htotal == 0)
+		drm_mode_set_crtcinfo(adjusted_mode, 0);
+
 	return true;
 }
 
drivers/gpu/drm/nouveau/nv50_instmem.c +2 -0
index 5f21df3..3fcae54 100644
@@ -141,6 +141,8 @@ nv50_instmem_init(struct drm_device *dev)
 	chan->file_priv = (struct drm_file *)-2;
 	dev_priv->fifos[0] = dev_priv->fifos[127] = chan;
 
+	INIT_LIST_HEAD(&chan->ramht_refs);
+
 	/* Channel's PRAMIN object + heap */
 	ret = nouveau_gpuobj_new_fake(dev, 0, c_offset, c_size, 0,
 							NULL, &chan->ramin);
drivers/message/fusion/mptscsih.c +2 -0
index 5c53624..407cb84 100644
@@ -2459,6 +2459,8 @@ mptscsih_slave_configure(struct scsi_device *sdev)
 		ioc->name,sdev->tagged_supported, sdev->simple_tags,
 		sdev->ordered_tags));
 
+	blk_queue_dma_alignment (sdev->request_queue, 512 - 1);
+
 	return 0;
 }
 
drivers/net/bonding/bond_3ad.c +3 -0
index 822f586..0ddf4c6 100644
@@ -2466,6 +2466,9 @@ int bond_3ad_lacpdu_recv(struct sk_buff *skb, struct net_device *dev, struct pac
 	if (!(dev->flags & IFF_MASTER))
 		goto out;
 
+	if (!pskb_may_pull(skb, sizeof(struct lacpdu)))
+		goto out;
+
 	read_lock(&bond->lock);
 	slave = bond_get_slave_by_dev((struct bonding *)netdev_priv(dev),
 					orig_dev);
drivers/net/bonding/bond_alb.c +3 -0
index 8d7dfd2..0cb1ded 100644
@@ -369,6 +369,9 @@ static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct
 		goto out;
 	}
 
+	if (!pskb_may_pull(skb, arp_hdr_len(bond_dev)))
+		goto out;
+
 	if (skb->len < sizeof(struct arp_pkt)) {
 		pr_debug("Packet is too small to be an ARP\n");
 		goto out;
drivers/net/cxgb3/cxgb3_main.c +2 -0
index e3f1b85..3e0f19f 100644
@@ -2296,6 +2296,8 @@ static int cxgb_extension_ioctl(struct net_device *dev, void __user *useraddr)
 	case CHELSIO_GET_QSET_NUM:{
 		struct ch_reg edata;
 
+		memset(&edata, 0, sizeof(struct ch_reg));
+
 		edata.cmd = CHELSIO_GET_QSET_NUM;
 		edata.val = pi->nqsets;
 		if (copy_to_user(useraddr, &edata, sizeof(edata)))
drivers/net/eql.c +2 -0
index dda2c79..0cb1cf9 100644
@@ -555,6 +555,8 @@ static int eql_g_master_cfg(struct net_device *dev, master_config_t __user *mcp)
 	equalizer_t *eql;
 	master_config_t mc;
 
+	memset(&mc, 0, sizeof(master_config_t));
+
 	if (eql_is_master(dev)) {
 		eql = netdev_priv(dev);
 		mc.max_slaves = eql->max_slaves;
drivers/net/usb/hso.c +2 -0
index 4dd2351..e254274 100644
@@ -1653,6 +1653,8 @@ static int hso_get_count(struct hso_serial *serial,
 	struct uart_icount cnow;
 	struct hso_tiocmget  *tiocmget = serial->tiocmget;
 
+	memset(&icount, 0, sizeof(struct serial_icounter_struct));
+
 	if (!tiocmget)
 		 return -ENOENT;
 	spin_lock_irq(&serial->serial_lock);
drivers/pci/intel-iommu.c +43 -47

108 lines not on this page · raw message

drivers/platform/x86/dell-wmi.c +1 -1
index 66f53c3..12a8e6f 100644
@@ -221,7 +221,7 @@ static void dell_wmi_notify(u32 value, void *context)
 			return;
 		}
 
-		if (dell_new_hk_type)
+		if (dell_new_hk_type || buffer_entry[1] == 0x0)
 			reported_key = (int)buffer_entry[2];
 		else
 			reported_key = (int)buffer_entry[1] & 0xffff;
drivers/platform/x86/hp-wmi.c +32 -19

82 lines not on this page · raw message

drivers/rtc/rtc-s3c.c +8 -5
index 70b68d3..2a305ef 100644
@@ -298,11 +298,6 @@ static int s3c_rtc_setalarm(struct device *dev, struct rtc_wkalrm *alrm)
 
 	s3c_rtc_setaie(alrm->enabled);
 
-	if (alrm->enabled)
-		enable_irq_wake(s3c_rtc_alarmno);
-	else
-		disable_irq_wake(s3c_rtc_alarmno);
-
 	return 0;
 }
 
@@ -547,6 +542,10 @@ static int s3c_rtc_suspend(struct platform_device *pdev, pm_message_t state)
 		ticnt_en_save &= S3C64XX_RTCCON_TICEN;
 	}
 	s3c_rtc_enable(pdev, 0);
+
+	if (device_may_wakeup(&pdev->dev))
+		enable_irq_wake(s3c_rtc_alarmno);
+
 	return 0;
 }
 
@@ -560,6 +559,10 @@ static int s3c_rtc_resume(struct platform_device *pdev)
 		tmp = readb(s3c_rtc_base + S3C2410_RTCCON);
 		writeb(tmp | ticnt_en_save, s3c_rtc_base + S3C2410_RTCCON);
 	}
+
+	if (device_may_wakeup(&pdev->dev))
+		disable_irq_wake(s3c_rtc_alarmno);
+
 	return 0;
 }
 #else
drivers/staging/vt6655/wpactl.c +8 -3
index 22c2fab..f79808e 100644
@@ -766,9 +766,14 @@ static int wpa_set_associate(PSDevice pDevice,
     DBG_PRT(MSG_LEVEL_DEBUG, KERN_INFO "wpa_ie_len = %d\n", param->u.wpa_associate.wpa_ie_len);
 
 
-	if (param->u.wpa_associate.wpa_ie &&
-	    copy_from_user(&abyWPAIE[0], param->u.wpa_associate.wpa_ie, param->u.wpa_associate.wpa_ie_len))
-	    return -EINVAL;
+	if (param->u.wpa_associate.wpa_ie_len) {
+		if (!param->u.wpa_associate.wpa_ie)
+			return -EINVAL;
+		if (param->u.wpa_associate.wpa_ie_len > sizeof(abyWPAIE))
+			return -EINVAL;
+		if (copy_from_user(&abyWPAIE[0], param->u.wpa_associate.wpa_ie, param->u.wpa_associate.wpa_ie_len))
+			return -EFAULT;
+	}
 
 	if (param->u.wpa_associate.mode == 1)
 	    pMgmt->eConfigMode = WMAC_CONFIG_IBSS_STA;
drivers/usb/musb/musb_debugfs.c +2 -3
index c79a5e3..9e8639d 100644
@@ -195,15 +195,14 @@ static const struct file_operations musb_regdump_fops = {
 
 static int musb_test_mode_open(struct inode *inode, struct file *file)
 {
-	file->private_data = inode->i_private;
-
 	return single_open(file, musb_test_mode_show, inode->i_private);
 }
 
 static ssize_t musb_test_mode_write(struct file *file,
 		const char __user *ubuf, size_t count, loff_t *ppos)
 {
-	struct musb		*musb = file->private_data;
+	struct seq_file		*s = file->private_data;
+	struct musb		*musb = s->private;
 	u8			test = 0;
 	char			buf[18];
 
drivers/usb/serial/mos7720.c +3 -0
index 30922a7..aa66581 100644
@@ -2024,6 +2024,9 @@ static int mos7720_ioctl(struct tty_struct *tty, struct file *file,
 
 	case TIOCGICOUNT:
 		cnow = mos7720_port->icount;
+
+		memset(&icount, 0, sizeof(struct serial_icounter_struct));
+
 		icount.cts = cnow.cts;
 		icount.dsr = cnow.dsr;
 		icount.rng = cnow.rng;
drivers/usb/serial/mos7840.c +3 -0
index 1c9b6e9..1a42bc2 100644
@@ -2285,6 +2285,9 @@ static int mos7840_ioctl(struct tty_struct *tty, struct file *file,
 	case TIOCGICOUNT:
 		cnow = mos7840_port->icount;
 		smp_rmb();
+
+		memset(&icount, 0, sizeof(struct serial_icounter_struct));
+
 		icount.cts = cnow.cts;
 		icount.dsr = cnow.dsr;
 		icount.rng = cnow.rng;
drivers/video/sis/sis_main.c +3 -0
index 559bf17..b52f8e4 100644
@@ -1701,6 +1701,9 @@ static int	sisfb_ioctl(struct fb_info *info, unsigned int cmd,
 		break;
 
 	   case FBIOGET_VBLANK:
+
+		memset(&sisvbblank, 0, sizeof(struct fb_vblank));
+
 		sisvbblank.count = 0;
 		sisvbblank.flags = sisfb_setupvbblankflags(ivideo, &sisvbblank.vcount, &sisvbblank.hcount);
 
drivers/video/via/ioctl.c +2 -0
index da03c07..4d553d0 100644
@@ -25,6 +25,8 @@ int viafb_ioctl_get_viafb_info(u_long arg)
 {
 	struct viafb_ioctl_info viainfo;
 
+	memset(&viainfo, 0, sizeof(struct viafb_ioctl_info));
+
 	viainfo.viafb_id = VIAID;
 	viainfo.vendor_id = PCI_VIA_VENDOR_ID;
 
fs/aio.c +12 -1
index 1ccf25c..5fb0fd7 100644
@@ -712,8 +712,16 @@ static ssize_t aio_run_iocb(struct kiocb *iocb)
 	 */
 	ret = retry(iocb);
 
-	if (ret != -EIOCBRETRY && ret != -EIOCBQUEUED)
+	if (ret != -EIOCBRETRY && ret != -EIOCBQUEUED) {
+		/*
+		 * There's no easy way to restart the syscall since other AIO's
+		 * may be already running. Just fail this IO with EINTR.
+		 */
+		if (unlikely(ret == -ERESTARTSYS || ret == -ERESTARTNOINTR ||
+			     ret == -ERESTARTNOHAND || ret == -ERESTART_RESTARTBLOCK))
+			ret = -EINTR;
 		aio_complete(iocb, ret, 0);
+	}
 out:
 	spin_lock_irq(&ctx->ctx_lock);
 
@@ -1659,6 +1667,9 @@ long do_io_submit(aio_context_t ctx_id, long nr,
 	if (unlikely(nr < 0))
 		return -EINVAL;
 
+	if (unlikely(nr > LONG_MAX/sizeof(*iocbpp)))
+		nr = LONG_MAX/sizeof(*iocbpp);
+
 	if (unlikely(!access_ok(VERIFY_READ, iocbpp, (nr*sizeof(*iocbpp)))))
 		return -EFAULT;
 
fs/char_dev.c +3 -1
index f80a4f2..143d393 100644
@@ -40,7 +40,9 @@ struct backing_dev_info directly_mappable_cdev_bdi = {
 #endif
 		/* permit direct mmap, for read, write or exec */
 		BDI_CAP_MAP_DIRECT |
-		BDI_CAP_READ_MAP | BDI_CAP_WRITE_MAP | BDI_CAP_EXEC_MAP),
+		BDI_CAP_READ_MAP | BDI_CAP_WRITE_MAP | BDI_CAP_EXEC_MAP |
+		/* no writeback happens */
+		BDI_CAP_NO_ACCT_AND_WRITEBACK),
 };
 
 static struct kobj_map *cdev_map;
fs/compat.c +1 -1
index 6490d21..af7c230 100644
@@ -1150,7 +1150,7 @@ static ssize_t compat_do_readv_writev(int type, struct file *file,
 {
 	compat_ssize_t tot_len;
 	struct iovec iovstack[UIO_FASTIOV];
-	struct iovec *iov;
+	struct iovec *iov = iovstack;
 	ssize_t ret;
 	io_fn_t fn;
 	iov_fn_t fnv;
fs/gfs2/log.c +1 -1
index 6a857e2..83917b5 100644
@@ -932,7 +932,7 @@ int gfs2_logd(void *data)
 
 		do {
 			prepare_to_wait(&sdp->sd_logd_waitq, &wait,
-					TASK_UNINTERRUPTIBLE);
+					TASK_INTERRUPTIBLE);
 			if (!gfs2_ail_flush_reqd(sdp) &&
 			    !gfs2_jrnl_flush_reqd(sdp) &&
 			    !kthread_should_stop())
fs/notify/inotify/inotify_user.c +5 -2
index e46ca68..0c6bbc0 100644
@@ -96,8 +96,11 @@ static inline __u32 inotify_arg_to_mask(u32 arg)
 {
 	__u32 mask;
 
-	/* everything should accept their own ignored and cares about children */
-	mask = (FS_IN_IGNORED | FS_EVENT_ON_CHILD);
+	/*
+	 * everything should accept their own ignored, cares about children,
+	 * and should receive events when the inode is unmounted
+	 */
+	mask = (FS_IN_IGNORED | FS_EVENT_ON_CHILD | FS_UNMOUNT);
 
 	/* mask off the flags used to open the fd */
 	mask |= (arg & (IN_ALL_EVENTS | IN_ONESHOT));
fs/proc/vmcore.c +1 -1
index 91c817f..2367fb3 100644
@@ -163,7 +163,7 @@ static ssize_t read_vmcore(struct file *file, char __user *buffer,
 
 static const struct file_operations proc_vmcore_operations = {
 	.read		= read_vmcore,
-	.llseek		= generic_file_llseek,
+	.llseek		= default_llseek,
 };
 
 static struct vmcore* __init get_new_element(void)
fs/xfs/linux-2.6/xfs_ioctl.c +2 -0
index e59a810..82a74f6 100644
@@ -794,6 +794,8 @@ xfs_ioc_fsgetxattr(
 {
 	struct fsxattr		fa;
 
+	memset(&fa, 0, sizeof(struct fsxattr));
+
 	xfs_ilock(ip, XFS_ILOCK_SHARED);
 	fa.fsx_xflags = xfs_ip2xflags(ip);
 	fa.fsx_extsize = ip->i_d.di_extsize << ip->i_mount->m_sb.sb_blocklog;
include/linux/ksm.h +9 -11
index 43bdab7..e809eb1 100644
@@ -16,6 +16,9 @@
 struct stable_node;
 struct mem_cgroup;
 
+struct page *ksm_does_need_to_copy(struct page *page,
+			struct vm_area_struct *vma, unsigned long address);
+
 #ifdef CONFIG_KSM
 int ksm_madvise(struct vm_area_struct *vma, unsigned long start,
 		unsigned long end, int advice, unsigned long *vm_flags);
@@ -70,19 +73,14 @@ static inline void set_page_stable_node(struct page *page,
  * We'd like to make this conditional on vma->vm_flags & VM_MERGEABLE,
  * but what if the vma was unmerged while the page was swapped out?
  */
-struct page *ksm_does_need_to_copy(struct page *page,
-			struct vm_area_struct *vma, unsigned long address);
-static inline struct page *ksm_might_need_to_copy(struct page *page,
+static inline int ksm_might_need_to_copy(struct page *page,
 			struct vm_area_struct *vma, unsigned long address)
 {
 	struct anon_vma *anon_vma = page_anon_vma(page);
 
-	if (!anon_vma ||
-	    (anon_vma == vma->anon_vma &&
-	     page->index == linear_page_index(vma, address)))
-		return page;
-
-	return ksm_does_need_to_copy(page, vma, address);
+	return anon_vma &&
+		(anon_vma != vma->anon_vma ||
+		 page->index != linear_page_index(vma, address));
 }
 
 int page_referenced_ksm(struct page *page,
@@ -115,10 +113,10 @@ static inline int ksm_madvise(struct vm_area_struct *vma, unsigned long start,
 	return 0;
 }
 
-static inline struct page *ksm_might_need_to_copy(struct page *page,
+static inline int ksm_might_need_to_copy(struct page *page,
 			struct vm_area_struct *vma, unsigned long address)
 {
-	return page;
+	return 0;
 }
 
 static inline int page_referenced_ksm(struct page *page,
include/linux/kvm_host.h +1 -0
index 7cb116a..f24d8fc 100644
@@ -123,6 +123,7 @@ struct kvm_memory_slot {
 	} *lpage_info[KVM_NR_PAGE_SIZES - 1];
 	unsigned long userspace_addr;
 	int user_alloc;
+	int id;
 };
 
 static inline unsigned long kvm_dirty_bitmap_bytes(struct kvm_memory_slot *memslot)
include/linux/mm.h +7 -1
index a2b4804..c1f54d2 100644
@@ -78,7 +78,11 @@ extern unsigned int kobjsize(const void *objp);
 #define VM_MAYSHARE	0x00000080
 
 #define VM_GROWSDOWN	0x00000100	/* general info on the segment */
+#if defined(CONFIG_STACK_GROWSUP) || defined(CONFIG_IA64)
 #define VM_GROWSUP	0x00000200
+#else
+#define VM_GROWSUP	0x00000000
+#endif
 #define VM_PFNMAP	0x00000400	/* Page-ranges managed without "struct page", just pure PFN */
 #define VM_DENYWRITE	0x00000800	/* ETXTBSY on write attempts.. */
 
@@ -1329,8 +1333,10 @@ unsigned long ra_submit(struct file_ra_state *ra,
 
 /* Do stack extension */
 extern int expand_stack(struct vm_area_struct *vma, unsigned long address);
-#ifdef CONFIG_IA64
+#if VM_GROWSUP
 extern int expand_upwards(struct vm_area_struct *vma, unsigned long address);
+#else
+  #define expand_upwards(vma, address) do { } while (0)
 #endif
 extern int expand_stack_downwards(struct vm_area_struct *vma,
 				  unsigned long address);
include/linux/mmzone.h +13 -0
index b4d109e..8b2db3d 100644
@@ -284,6 +284,13 @@ struct zone {
 	unsigned long watermark[NR_WMARK];
 
 	/*
+	 * When free pages are below this point, additional steps are taken
+	 * when reading the number of free pages to avoid per-cpu counter
+	 * drift allowing watermarks to be breached
+	 */
+	unsigned long percpu_drift_mark;
+
+	/*
 	 * We don't know if the memory that we're going to allocate will be freeable
 	 * or/and it will be released eventually, so to avoid totally wasting several
 	 * GB of ram we must reserve some of the lower zone memory (otherwise we risk
@@ -456,6 +463,12 @@ static inline int zone_is_oom_locked(const struct zone *zone)
 	return test_bit(ZONE_OOM_LOCKED, &zone->flags);
 }
 
+#ifdef CONFIG_SMP
+unsigned long zone_nr_free_pages(struct zone *zone);
+#else
+#define zone_nr_free_pages(zone) zone_page_state(zone, NR_FREE_PAGES)
+#endif /* CONFIG_SMP */
+
 /*
  * The "priority" of VM scanning is how much of the queues we will scan in one
  * go. A value of 12 for DEF_PRIORITY implies that we will scan 1/4096th of the
include/linux/vmstat.h +22 -0
index 7f43ccd..eaaea37 100644
@@ -170,6 +170,28 @@ static inline unsigned long zone_page_state(struct zone *zone,
 	return x;
 }
 
+/*
+ * More accurate version that also considers the currently pending
+ * deltas. For that we need to loop over all cpus to find the current
+ * deltas. There is no synchronization so the result cannot be
+ * exactly accurate either.
+ */
+static inline unsigned long zone_page_state_snapshot(struct zone *zone,
+					enum zone_stat_item item)
+{
+	long x = atomic_long_read(&zone->vm_stat[item]);
+
+#ifdef CONFIG_SMP
+	int cpu;
+	for_each_online_cpu(cpu)
+		x += per_cpu_ptr(zone->pageset, cpu)->vm_stat_diff[item];
+
+	if (x < 0)
+		x = 0;
+#endif
+	return x;
+}
+
 extern unsigned long global_reclaimable_pages(void);
 extern unsigned long zone_reclaimable_pages(struct zone *zone);
 
include/net/sock.h +1 -0
index 0a691ea..e2a9f59 100644
@@ -749,6 +749,7 @@ struct proto {
 	/* Keeping track of sk's, looking them up, and port selection methods. */
 	void			(*hash)(struct sock *sk);
 	void			(*unhash)(struct sock *sk);
+	void			(*rehash)(struct sock *sk);
 	int			(*get_port)(struct sock *sk, unsigned short snum);
 
 	/* Keeping track of sockets in use */
include/net/tcp.h +30 -6
index a144914..e824cf0 100644
@@ -268,11 +268,21 @@ static inline int between(__u32 seq1, __u32 seq2, __u32 seq3)
 	return seq3 - seq2 >= seq1 - seq2;
 }
 
-static inline int tcp_too_many_orphans(struct sock *sk, int num)
+static inline bool tcp_too_many_orphans(struct sock *sk, int shift)
 {
-	return (num > sysctl_tcp_max_orphans) ||
-		(sk->sk_wmem_queued > SOCK_MIN_SNDBUF &&
-		 atomic_read(&tcp_memory_allocated) > sysctl_tcp_mem[2]);
+	struct percpu_counter *ocp = sk->sk_prot->orphan_count;
+	int orphans = percpu_counter_read_positive(ocp);
+
+	if (orphans << shift > sysctl_tcp_max_orphans) {
+		orphans = percpu_counter_sum_positive(ocp);
+		if (orphans << shift > sysctl_tcp_max_orphans)
+			return true;
+	}
+
+	if (sk->sk_wmem_queued > SOCK_MIN_SNDBUF &&
+	    atomic_read(&tcp_memory_allocated) > sysctl_tcp_mem[2])
+		return true;
+	return false;
 }
 
 /* syncookies: remember time of last synqueue overflow */
@@ -509,8 +519,22 @@ extern unsigned int tcp_current_mss(struct sock *sk);
 /* Bound MSS / TSO packet size with the half of the window */
 static inline int tcp_bound_to_half_wnd(struct tcp_sock *tp, int pktsize)
 {
-	if (tp->max_window && pktsize > (tp->max_window >> 1))
-		return max(tp->max_window >> 1, 68U - tp->tcp_header_len);
+	int cutoff;
+
+	/* When peer uses tiny windows, there is no use in packetizing
+	 * to sub-MSS pieces for the sake of SWS or making sure there
+	 * are enough packets in the pipe for fast recovery.
+	 *
+	 * On the other hand, for extremely large MSS devices, handling
+	 * smaller than MSS windows in this way does make sense.
+	 */
+	if (tp->max_window >= 512)
+		cutoff = (tp->max_window >> 1);
+	else
+		cutoff = tp->max_window;
+
+	if (cutoff && pktsize > cutoff)
+		return max_t(int, cutoff, 68U - tp->tcp_header_len);
 	else
 		return pktsize;
 }
include/net/udp.h +1 -0
index 5348d80..4201dc8 100644
@@ -151,6 +151,7 @@ static inline void udp_lib_hash(struct sock *sk)
 }
 
 extern void udp_lib_unhash(struct sock *sk);
+extern void udp_lib_rehash(struct sock *sk, u16 new_hash);
 
 static inline void udp_lib_close(struct sock *sk, long timeout)
 {
kernel/exit.c +2 -3
index ceffc67..ac90425 100644
@@ -1383,8 +1383,7 @@ static int wait_task_stopped(struct wait_opts *wo,
 	if (!unlikely(wo->wo_flags & WNOWAIT))
 		*p_code = 0;
 
-	/* don't need the RCU readlock here as we're holding a spinlock */
-	uid = __task_cred(p)->uid;
+	uid = task_uid(p);
 unlock_sig:
 	spin_unlock_irq(&p->sighand->siglock);
 	if (!exit_code)
@@ -1457,7 +1456,7 @@ static int wait_task_continued(struct wait_opts *wo, struct task_struct *p)
 	}
 	if (!unlikely(wo->wo_flags & WNOWAIT))
 		p->signal->flags &= ~SIGNAL_STOP_CONTINUED;
-	uid = __task_cred(p)->uid;
+	uid = task_uid(p);
 	spin_unlock_irq(&p->sighand->siglock);
 
 	pid = task_pid_vnr(p);
kernel/hw_breakpoint.c +2 -1
index 7a56b22..6eb6422 100644
@@ -417,7 +417,8 @@ register_user_hw_breakpoint(struct perf_event_attr *attr,
 			    perf_overflow_handler_t triggered,
 			    struct task_struct *tsk)
 {
-	return perf_event_create_kernel_counter(attr, -1, tsk->pid, triggered);
+	return perf_event_create_kernel_counter(attr, -1, task_pid_vnr(tsk),
+						triggered);
 }
 EXPORT_SYMBOL_GPL(register_user_hw_breakpoint);
 
kernel/power/snapshot.c +65 -20

130 lines not on this page · raw message

kernel/sched.c +4 -4
index 6d0dbeb..4fa520b 100644
@@ -3349,9 +3349,9 @@ void task_times(struct task_struct *p, cputime_t *ut, cputime_t *st)
 	rtime = nsecs_to_cputime(p->se.sum_exec_runtime);
 
 	if (total) {
-		u64 temp;
+		u64 temp = rtime;
 
-		temp = (u64)(rtime * utime);
+		temp *= utime;
 		do_div(temp, total);
 		utime = (cputime_t)temp;
 	} else
@@ -3382,9 +3382,9 @@ void thread_group_times(struct task_struct *p, cputime_t *ut, cputime_t *st)
 	rtime = nsecs_to_cputime(cputime.sum_exec_runtime);
 
 	if (total) {
-		u64 temp;
+		u64 temp = rtime;
 
-		temp = (u64)(rtime * cputime.utime);
+		temp *= cputime.utime;
 		do_div(temp, total);
 		utime = (cputime_t)temp;
 	} else
kernel/sys.c +2 -0
index e83ddbb..f4fcc6e 100644
@@ -931,6 +931,7 @@ SYSCALL_DEFINE2(setpgid, pid_t, pid, pid_t, pgid)
 		pgid = pid;
 	if (pgid < 0)
 		return -EINVAL;
+	rcu_read_lock();
 
 	/* From this point forward we keep holding onto the tasklist lock
 	 * so that our parent does not change from under us. -DaveM
@@ -984,6 +985,7 @@ SYSCALL_DEFINE2(setpgid, pid_t, pid, pid_t, pgid)
 out:
 	/* All paths lead to here, thus we are safe. -DaveM */
 	write_unlock_irq(&tasklist_lock);
+	rcu_read_unlock();
 	return err;
 }
 
mm/backing-dev.c +2 -0
index 123bcef..c0f7157 100644
@@ -29,6 +29,7 @@ EXPORT_SYMBOL_GPL(default_backing_dev_info);
 
 struct backing_dev_info noop_backing_dev_info = {
 	.name		= "noop",
+	.capabilities	= BDI_CAP_NO_ACCT_AND_WRITEBACK,
 };
 EXPORT_SYMBOL_GPL(noop_backing_dev_info);
 
@@ -254,6 +255,7 @@ static int __init default_bdi_init(void)
 	err = bdi_init(&default_backing_dev_info);
 	if (!err)
 		bdi_register(&default_backing_dev_info, NULL, "default");
+	err = bdi_init(&noop_backing_dev_info);
 
 	return err;
 }
mm/ksm.c +0 -3
index 6c3e99b..17ee9df 100644
@@ -1523,8 +1523,6 @@ struct page *ksm_does_need_to_copy(struct page *page,
 {
 	struct page *new_page;
 
-	unlock_page(page);	/* any racers will COW it, not modify it */
-
 	new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address);
 	if (new_page) {
 		copy_user_highpage(new_page, page, address, vma);
@@ -1540,7 +1538,6 @@ struct page *ksm_does_need_to_copy(struct page *page,
 			add_page_to_unevictable_list(new_page);
 	}
 
-	page_cache_release(page);
 	return new_page;
 }
 
mm/memory.c +47 -9

91 lines not on this page · raw message

mm/mmap.c +1 -3
index 3867cfc..3a94362 100644
@@ -1704,9 +1704,6 @@ static int acct_stack_growth(struct vm_area_struct *vma, unsigned long size, uns
  * PA-RISC uses this for its stack; IA64 for its Register Backing Store.
  * vma is the last one with address > vma->vm_end.  Have to extend vma.
  */
-#ifndef CONFIG_IA64
-static
-#endif
 int expand_upwards(struct vm_area_struct *vma, unsigned long address)
 {
 	int error;
@@ -1997,6 +1994,7 @@ static int __split_vma(struct mm_struct * mm, struct vm_area_struct * vma,
 			removed_exe_file_vma(mm);
 		fput(new->vm_file);
 	}
+	unlink_anon_vmas(new);
  out_free_mpol:
 	mpol_put(pol);
  out_free_vma:
mm/mmzone.c +21 -0
index f5b7d17..e35bfb8 100644
@@ -87,3 +87,24 @@ int memmap_valid_within(unsigned long pfn,
 	return 1;
 }
 #endif /* CONFIG_ARCH_HAS_HOLES_MEMORYMODEL */
+
+#ifdef CONFIG_SMP
+/* Called when a more accurate view of NR_FREE_PAGES is needed */
+unsigned long zone_nr_free_pages(struct zone *zone)
+{
+	unsigned long nr_free_pages = zone_page_state(zone, NR_FREE_PAGES);
+
+	/*
+	 * While kswapd is awake, it is considered the zone is under some
+	 * memory pressure. Under pressure, there is a risk that
+	 * per-cpu-counter-drift will allow the min watermark to be breached
+	 * potentially causing a live-lock. While kswapd is awake and
+	 * free pages are low, get a better estimate for free pages
+	 */
+	if (nr_free_pages < zone->percpu_drift_mark &&
+			!waitqueue_active(&zone->zone_pgdat->kswapd_wait))
+		return zone_page_state_snapshot(zone, NR_FREE_PAGES);
+
+	return nr_free_pages;
+}
+#endif /* CONFIG_SMP */
mm/page_alloc.c +23 -10
index 9bd339e..f7cc624 100644
@@ -588,13 +588,13 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 {
 	int migratetype = 0;
 	int batch_free = 0;
+	int to_free = count;
 
 	spin_lock(&zone->lock);
 	zone->all_unreclaimable = 0;
 	zone->pages_scanned = 0;
 
-	__mod_zone_page_state(zone, NR_FREE_PAGES, count);
-	while (count) {
+	while (to_free) {
 		struct page *page;
 		struct list_head *list;
 
@@ -619,8 +619,9 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 			/* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */
 			__free_one_page(page, zone, 0, page_private(page));
 			trace_mm_page_pcpu_drain(page, 0, page_private(page));
-		} while (--count && --batch_free && !list_empty(list));
+		} while (--to_free && --batch_free && !list_empty(list));
 	}
+	__mod_zone_page_state(zone, NR_FREE_PAGES, count);
 	spin_unlock(&zone->lock);
 }
 
@@ -631,8 +632,8 @@ static void free_one_page(struct zone *zone, struct page *page, int order,
 	zone->all_unreclaimable = 0;
 	zone->pages_scanned = 0;
 
-	__mod_zone_page_state(zone, NR_FREE_PAGES, 1 << order);
 	__free_one_page(page, zone, order, migratetype);
+	__mod_zone_page_state(zone, NR_FREE_PAGES, 1 << order);
 	spin_unlock(&zone->lock);
 }
 
@@ -1461,7 +1462,7 @@ int zone_watermark_ok(struct zone *z, int order, unsigned long mark,
 {
 	/* free_pages my go negative - that's OK */
 	long min = mark;
-	long free_pages = zone_page_state(z, NR_FREE_PAGES) - (1 << order) + 1;
+	long free_pages = zone_nr_free_pages(z) - (1 << order) + 1;
 	int o;
 
 	if (alloc_flags & ALLOC_HIGH)
@@ -1843,6 +1844,7 @@ __alloc_pages_direct_reclaim(gfp_t gfp_mask, unsigned int order,
 	struct page *page = NULL;
 	struct reclaim_state reclaim_state;
 	struct task_struct *p = current;
+	bool drained = false;
 
 	cond_resched();
 
@@ -1861,14 +1863,25 @@ __alloc_pages_direct_reclaim(gfp_t gfp_mask, unsigned int order,
 
 	cond_resched();
 
-	if (order != 0)
-		drain_all_pages();
+	if (unlikely(!(*did_some_progress)))
+		return NULL;
 
-	if (likely(*did_some_progress))
-		page = get_page_from_freelist(gfp_mask, nodemask, order,
+retry:
+	page = get_page_from_freelist(gfp_mask, nodemask, order,
 					zonelist, high_zoneidx,
 					alloc_flags, preferred_zone,
 					migratetype);
+
+	/*
+	 * If an allocation failed after direct reclaim, it could be because
+	 * pages are pinned on the per-cpu lists. Drain them and try again
+	 */
+	if (!page && !drained) {
+		drain_all_pages();
+		drained = true;
+		goto retry;
+	}
+
 	return page;
 }
 
@@ -2412,7 +2425,7 @@ void show_free_areas(void)
 			" all_unreclaimable? %s"
 			"\n",
 			zone->name,
-			K(zone_page_state(zone, NR_FREE_PAGES)),
+			K(zone_nr_free_pages(zone)),
 			K(min_wmark_pages(zone)),
 			K(low_wmark_pages(zone)),
 			K(high_wmark_pages(zone)),
mm/percpu.c +1 -1
index 6470e77..0a09fff 100644
@@ -1413,9 +1413,9 @@ int __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai,
 
 			if (pcpu_first_unit_cpu == NR_CPUS)
 				pcpu_first_unit_cpu = cpu;
+			pcpu_last_unit_cpu = cpu;
 		}
 	}
-	pcpu_last_unit_cpu = cpu;
 	pcpu_nr_units = unit;
 
 	for_each_possible_cpu(cpu)
mm/vmscan.c +35 -8

85 lines not on this page · raw message

mm/vmstat.c +14 -1
index 7759941..26d5716 100644
@@ -138,11 +138,24 @@ static void refresh_zone_stat_thresholds(void)
 	int threshold;
 
 	for_each_populated_zone(zone) {
+		unsigned long max_drift, tolerate_drift;
+
 		threshold = calculate_threshold(zone);
 
 		for_each_online_cpu(cpu)
 			per_cpu_ptr(zone->pageset, cpu)->stat_threshold
 							= threshold;
+
+		/*
+		 * Only set percpu_drift_mark if there is a danger that
+		 * NR_FREE_PAGES reports the low watermark is ok when in fact
+		 * the min watermark could be breached by an allocation
+		 */
+		tolerate_drift = low_wmark_pages(zone) - min_wmark_pages(zone);
+		max_drift = num_online_cpus() * threshold;
+		if (max_drift > tolerate_drift)
+			zone->percpu_drift_mark = high_wmark_pages(zone) +
+					max_drift;
 	}
 }
 
@@ -813,7 +826,7 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
 		   "\n        scanned  %lu"
 		   "\n        spanned  %lu"
 		   "\n        present  %lu",
-		   zone_page_state(zone, NR_FREE_PAGES),
+		   zone_nr_free_pages(zone),
 		   min_wmark_pages(zone),
 		   low_wmark_pages(zone),
 		   high_wmark_pages(zone),
net/Kconfig +1 -1
index 0d68b40..a44092e 100644
@@ -206,7 +206,7 @@ source "net/dcb/Kconfig"
 
 config RPS
[… more lines in the raw message]

Lines of this file are not shown. View the raw message.

net/bridge/br_netfilter.c +4 -2
index 8fb75f8..e759b89 100644

Lines of this file are not shown. View the raw message.

net/core/skbuff.c +5 -1
index ce88293..a586e89 100644

Lines of this file are not shown. View the raw message.

net/ipv4/datagram.c +4 -1
index fb24658..31cafd4 100644

Lines of this file are not shown. View the raw message.

net/ipv4/route.c +6 -1
index 560acc6..0189deb 100644

Lines of this file are not shown. View the raw message.

net/ipv4/tcp.c +10 -22
index c259714..c88ca26 100644

Lines of this file are not shown. View the raw message.

net/ipv4/tcp_timer.c +4 -4
index 440a5c6..7abd922 100644

Lines of this file are not shown. View the raw message.

net/ipv4/udp.c +44 -0

44 lines not on this page · raw message

net/ipv6/datagram.c +6 -1
index 7126846..0e3754a1 100644

Lines of this file are not shown. View the raw message.

net/ipv6/udp.c +10 -0
index 87be586..97e1214 100644

Lines of this file are not shown. View the raw message.

net/irda/af_irda.c +2 -2
index 79986a6..fd55b51 100644

Lines of this file are not shown. View the raw message.

net/l2tp/l2tp_eth.c +1 -1
index 58c6c4c..1ae6976 100644

Lines of this file are not shown. View the raw message.

net/llc/af_llc.c +2 -1
index 023ba82..5826129 100644

Lines of this file are not shown. View the raw message.

net/rds/recv.c +1 -1
index 795a00b..c93588c 100644

Lines of this file are not shown. View the raw message.

net/sctp/output.c +0 -1
index a646681..bcc4590 100644

Lines of this file are not shown. View the raw message.

net/unix/af_unix.c +12 -3
index fef2cc5..617bea4 100644

Lines of this file are not shown. View the raw message.

security/keys/keyctl.c +5 -1
index 6261745..e0d1e2b 100644

Lines of this file are not shown. View the raw message.

sound/pci/hda/patch_sigmatel.c +11 -1
index f1e7bab..b8d730c 100644

Lines of this file are not shown. View the raw message.

virt/kvm/kvm_main.c +1 -0
index f032806..2774528 100644

Lines of this file are not shown. View the raw message.

The rest of this diff is not shown. View the raw message.

Part of the diff is not shown here. View the raw message.

#2 lore raw
In reply to Greg KH

Hi,

Quoted 8 lines · Greg KH

I'm announcing the release of the 2.6.35.6 kernel.

All users of the 2.6.35 kernel series must upgrade.

The updated 2.6.35.y git tree can be found at:
       git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.35.y.git
and can be browsed at the normal kernel.org git web browser:
       http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.35.y.git;a=summary

I usually do 'git fetch' to get new changes in all my repos, however, this one seems to be a bit different; there's no branch. Since there's no branch, 'git fetch' doesn't retrieve the tags.

Can you add a 'master' branch that points to the latest tag? That would make things easier for other people, specially the ones that just do 'git pull'.

Cheers.

Signature · 1 line
Felipe Contreras
author
#3 lore raw
Quoted 12 lines · Felipe Contreras

Hi,

Quoted 8 lines · Greg KH

I'm announcing the release of the 2.6.35.6 kernel.

All users of the 2.6.35 kernel series must upgrade.

The updated 2.6.35.y git tree can be found at:
?? ?? ?? ??git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.35.y.git and can be browsed at the normal kernel.org git web browser:
?? ?? ?? ??http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.35.y.git;a=summary

Quoted 7 lines

I usually do 'git fetch' to get new changes in all my repos, however, this one seems to be a bit different; there's no branch. Since there's no branch, 'git fetch' doesn't retrieve the tags.

Can you add a 'master' branch that points to the latest tag? That would make things easier for other people, specially the ones that just do 'git pull'.

I don't understand, this works for me this way, the master branch _is_ the latest version at all times. Did you just need to wait a few minutes to have the mirror catch up?

thanks,

greg k-h

#5 lore raw
In reply to Greg KH

I'm not exactly sure what's going on here and I'd like some help
figuring out what is. For some inexplicable reason, ever since I started using the 2.6.35.x series with the 2.6.35.3 release, my loadaverages tend to bounce anywhere from as low to .2 to 1.5 - constantly, while the machine seems idle. Consistently, there are no programs in D state in ps aux output, no cpu hogging programs running in top, nothing I can see that should explain the bizzarely high load average. *Something* is wrong beyond the mere loadaverage numbers going crazy however, since timed runs of kernel compiles done with my distro's kernel and 2.6.35.5 show that while there is no *apparent* use of cpu or disk showing in vmstat while the machine is idle, the compiles on the newer kernel are taking approximately twice as long as before. Now, while I could try to figure out what patch started the problem, I think it would be a better idea for me to make sure I'm looking for the actual problem in the right place. Therefore, I know that cpu use, disk I/O, and the kernel can drive the load average up. Of these things, cpu use and disk I/O are trackable in top/ps output (eg, a process in D state is waiting on the disk to do something and can't sleep.) but I don't know if there's any easy way to determine if the kernel itself is doing something that's driving up the load average. It's perplexing me that I can see the loadaverage constantly bouncing about but can't seem to find any reason why it is doing so.

If you have any tips or recommendations on what I should use to
investigate this further, please let me know. Once I have ensured to my own satisfaction that I'm not doing something bizzare that's screwing up my machine, I'll make a detailed bug report and start on figuring out how to use git bisect.

Tim McGrath

#4 lore raw
In reply to Greg KH
Quoted 8 lines · Greg KH
Quoted 3 lines · Felipe Contreras

Can you add a 'master' branch that points to the latest tag? That would make things easier for other people, specially the ones that just do 'git pull'.

I don't understand, this works for me this way, the master branch _is_ the latest version at all times.  Did you just need to wait a few minutes to have the mirror catch up?

It wasn't about timing: 'git branch -a' didn't show any remote branch. Anyway, I tried cloning from scratch, and cloning with a template and the branch was there. Also, I tried removing and creating the remote again and voila; the branch is not there. But have been a weird git bug.

Sorry for the noise.

Signature · 1 line
Felipe Contreras
author
#6 lore raw
In reply to [email protected]
Quoted 5 lines · On Mon, Sep 27, 2010 at 12:32:08PM -0400, [email protected] wrote:

I'm not exactly sure what's going on here and I'd like some help
figuring out what is. For some inexplicable reason, ever since I started using the 2.6.35.x series with the 2.6.35.3 release, my loadaverages tend to bounce anywhere from as low to .2 to 1.5 - constantly, while the machine seems idle.

But .34 doesn't show this issue?

If so, can you run 'git bisect' between the .34 and .35 releases to find out where the problem comes in?

Quoted 3 lines

Consistently, there are no programs in D state in ps aux output, no cpu hogging programs running in top, nothing I can see that should explain the bizzarely high load average.

Loadaverage can be affected by a driver doing an uninteruptable sleep. This doesn't actually cause a load on the system, but it does affect the way that number is calcuated. Perhaps that's the problem here.

Quoted 5 lines

If you have any tips or recommendations on what I should use to
investigate this further, please let me know. Once I have ensured to my own satisfaction that I'm not doing something bizzare that's screwing up my machine, I'll make a detailed bug report and start on figuring out how to use git bisect.

'man git-bisect' will help with this :)

thanks,

greg k-h

#7 lore raw
Quoted 11 lines · Greg KH
Quoted 5 lines · On Mon, Sep 27, 2010 at 12:32:08PM -0400, [email protected] wrote:

I'm not exactly sure what's going on here and I'd like some help
figuring out what is. For some inexplicable reason, ever since I started using the 2.6.35.x series with the 2.6.35.3 release, my loadaverages tend to bounce anywhere from as low to .2 to 1.5 - constantly, while the machine seems idle.

But .34 doesn't show this issue?

If so, can you run 'git bisect' between the .34 and .35 releases to find out where the problem comes in?

.34 does not, that is correct.

Quoted 9 lines
Quoted 4 lines

Consistently, there are no programs in D state in ps aux output, no cpu hogging programs running in top, nothing I can see that should explain the bizzarely high load average.

Loadaverage can be affected by a driver doing an uninteruptable sleep. This doesn't actually cause a load on the system, but it does affect the way that number is calcuated. Perhaps that's the problem here.

That would explain a lot.

Quoted 11 lines
Quoted 5 lines

If you have any tips or recommendations on what I should use to
investigate this further, please let me know. Once I have ensured to my own satisfaction that I'm not doing something bizzare that's screwing up my machine, I'll make a detailed bug report and start on figuring out how to use git bisect.

'man git-bisect' will help with this :)

thanks,

greg k-h

Thank you for the tips. If I don't find anything obvious I'll bisect it very soon, but I'll likely attempt to track backwards from .3 myself since I've backtracked through patches before. Once I have narrowed it down to a patch release it'll likely take less time to bisect. (eg, if I know it happened between 2.6.35-rc1 and rc2 I'd only have to bisect the patches between those versions, whereas right now I only know positively between 2.6.34 and 2.6.35.3 something went wrong.)

Tim McGrath

#8 lore raw
In reply to [email protected]
Quoted 13 lines · On Mon, 27 Sep 2010 12:32:08 -0400 [email protected] wrote:

I'm not exactly sure what's going on here and I'd like some help
figuring out what is. For some inexplicable reason, ever since I started using the 2.6.35.x series with the 2.6.35.3 release, my loadaverages tend to bounce anywhere from as low to .2 to 1.5 - constantly, while the machine seems idle. Consistently, there are no programs in D state in ps aux output, no cpu hogging programs running in top, nothing I can see that should explain the bizzarely high load average. *Something* is wrong beyond the mere loadaverage numbers going crazy however, since timed runs of kernel compiles done with my distro's kernel and 2.6.35.5 show that while there is no *apparent* use of cpu or disk showing in vmstat while the machine is idle, the compiles on the newer kernel are taking approximately twice as long as before. Now, while I could try to figure out what patch started the problem, I think it would be a better idea for me to make sure I'm looking

Quoted 8 lines

for the actual problem in the right place. Therefore, I know that cpu use, disk I/O, and the kernel can drive the load average up. Of these things, cpu use and disk I/O are trackable in top/ps output (eg, a process in D state is waiting on the disk to do something and can't sleep.) but I don't know if there's any easy way to determine if the kernel itself is doing something that's driving up the load average. It's perplexing me that I can see the loadaverage constantly bouncing about but can't seem to find any reason why it is doing so.

I can't comment on the bouncing part, but increased load average is partly bisected down in Bug 16525 (unexpected high load since 2.6.35 Bug 16525)[1]. Maybe you are seeing something similiar?

Quoted 9 lines

If you have any tips or recommendations on what I should use to
investigate this further, please let me know. Once I have ensured to my own satisfaction that I'm not doing something bizzare that's screwing up my machine, I'll make a detailed bug report and start on figuring out how to use git bisect.

Tim McGrath

#9 lore raw
Quoted 3 lines · Florian Mickler

I can't comment on the bouncing part, but increased load average is partly bisected down in Bug 16525 (unexpected high load since 2.6.35 Bug 16525)[1]. Maybe you are seeing something similiar?

If you're referring to

https://bugzilla.kernel.org/show_bug.cgi?id=16525

then I have to admit that's very interesting; in fact, I tested my machine without X running, and the problem I'm having goes away entirely - loadavg quickly settles to 0.00 - and even with *only* a mrxvt terminal running, that is to say NO window manager, just mrxvt running, X on 2.6.35.3 and later has the wildly bouncy load averages.

Here's a graphical example of just how wacky this is:

http://yfrog.com/6lloadbp

In this image, the dip down to less than 0.5 after the 18'th is due to me experimenting using the slackware distribution kernel (2.6.33.4) after I finally noticed something was amiss. The sharp rise afterwards is due to me first, building 2.6.35.5, and then afterwards, using it. To be perfectly clear, I've previously used 2.6.34.2 and did not experience the problem there either, nor is it in 2.6.33.4.

The majority of time during the last month the machine has sat idle, with nothing to do except idle with the screen turned off. Besides building kernels, the most difficult task I've asked of it has been websurfing. This is not exactly a difficult job for this machine, and hardly would normally push the load average up that high for that long.

Greg, I noticed that in the ChangeLogs for 2.6.35.x there were some patches for the intel driver which seem to be to my untrained eye an attempt to fix Linus' problem talked about in that bug report; The patch that is referenced in the bug report contains some things at least that were not applied to the tree; since I have no idea what I'm looking at, could you verify for me if I'm barking up the right tree here? If I am, I'll try building a clean 2.6.35 kernel with that patch applied and see if my problem disappears.

Anyway, I am now going to start going backwards through 2.6.35.x until I discover which version(s) caused my particular issue. Since 2.6.35.3 and later have had the problem, and 2.6.35.2 will not compile for me, I'm going to first compile and test 2.6.35.1.

This will probably take some time as the bouncy load averages are consistently inconsistent and tend to, much like a watched kettle, not show up when I expect them to, then do so immediately when I turn my back. :)

Thanks for the tip,
Tim McGrath

#10 lore raw
Quoted 7 lines · On Mon, Sep 27, 2010 at 07:39:56PM -0400, [email protected] wrote:

Greg, I noticed that in the ChangeLogs for 2.6.35.x there were some patches for the intel driver which seem to be to my untrained eye an attempt to fix Linus' problem talked about in that bug report; The patch that is referenced in the bug report contains some things at least that were not applied to the tree; since I have no idea what I'm looking at, could you verify for me if I'm barking up the right tree here? If I am, I'll try building a clean 2.6.35 kernel with that patch applied and see if my problem disappears.

WRT to this patch, it appears I was indeed barking up the wrong tree - it doesn't apply to 2.6.35 at all. I guess I'll have to start bisecting now. 2.6.35.1 and 2.6.35 clearly have the problem. Just wanted to make it clear that I'd been mistaken about this before you wasted time digging into it.

Thank you,
Tim McGrath

#11 lore raw
In reply to [email protected]
Quoted 7 lines · On Mon, 27 Sep 2010 19:39:56 -0400 [email protected] wrote:
Quoted 3 lines · Florian Mickler

I can't comment on the bouncing part, but increased load average is partly bisected down in Bug 16525 (unexpected high load since 2.6.35 Bug 16525)[1]. Maybe you are seeing something similiar?

If you're referring to

Quoted 8 lines

https://bugzilla.kernel.org/show_bug.cgi?id=16525

then I have to admit that's very interesting; in fact, I tested my machine without X running, and the problem I'm having goes away entirely - loadavg quickly settles to 0.00 - and even with *only* a mrxvt terminal running, that is to say NO window manager, just mrxvt running, X on 2.6.35.3 and later has the wildly bouncy load averages.

That would of course lead one to think it happens
somewhere in the gfx stack.

Quoted 11 lines

Here's a graphical example of just how wacky this is:

http://yfrog.com/6lloadbp

In this image, the dip down to less than 0.5 after the 18'th is due to me experimenting using the slackware distribution kernel (2.6.33.4) after I finally noticed something was amiss. The sharp rise afterwards is due to me first, building 2.6.35.5, and then afterwards, using it. To be perfectly clear, I've previously used 2.6.34.2 and did not experience the problem there either, nor is it in 2.6.33.4.

What load figure are you basing your observations on? The 15 minutes average should be the most interesting (sampled at a 7 minutes interval...)

Regards,
Flo

#12 lore raw
Quoted 12 lines · Florian Mickler
Quoted 11 lines

Here's a graphical example of just how wacky this is:

http://yfrog.com/6lloadbp

In this image, the dip down to less than 0.5 after the 18'th is due to me experimenting using the slackware distribution kernel (2.6.33.4) after I finally noticed something was amiss. The sharp rise afterwards is due to me first, building 2.6.35.5, and then afterwards, using it. To be perfectly clear, I've previously used 2.6.34.2 and did not experience the problem there either, nor is it in 2.6.33.4.

Quoted 3 lines

What load figure are you basing your observations on? The 15 minutes average should be the most interesting (sampled at a 7 minutes interval...)

my observations are based on letting the machine idle immediately after bootup. I monitor the state of the machine using a program called conky, which I have configured to show disk I/O, cpu use, swap I/O and among other things, the load average. Immediately after booting my loadaverage tends to peak at about 2.5 to 3.0; on a working kernel this eventually settles down to 0.00 to 0.05 in about ten minutes. On kernels that exhibit this problem, it doesn't settle lower than 0.3 and is much more likely to hang anywhere from 0.8 to 1.2. In fact, if I give it enough time it'll raise and lower itself constantly without any (visible) work being done. So basically I boot the machine and go get a drink, come back, and if it's been ten minutes, there's been no disk IO, cpu use, or any other activity recorded and it's still above 0.3 something's not working right.

At least, I think that was your question. If you're talking about the graph, I merely posted it to show that I've been having this problem for over a month, and it's demonstrably causing very inconsistent load averages. (Which is why the graph isn't anything close to a line, it's a mess!) the graph takes a reading every five minutes, if you were wondering about the sample rate.

If that didn't answer your question I didn't understand it. :)

In other news, I'm in the process of bisection but keep having to skip bisects that have compile errors. sigh. still at 12 hops, somewhere around five thousands commits to check.

Tim McGrath

#13 lore raw
Quoted 17 lines · On Tue, 28 Sep 2010 15:03:58 -0400 [email protected] wrote:
Quoted 12 lines · Florian Mickler
Quoted 11 lines

Here's a graphical example of just how wacky this is:

http://yfrog.com/6lloadbp

In this image, the dip down to less than 0.5 after the 18'th is due to me experimenting using the slackware distribution kernel (2.6.33.4) after I finally noticed something was amiss. The sharp rise afterwards is due to me first, building 2.6.35.5, and then afterwards, using it. To be perfectly clear, I've previously used 2.6.34.2 and did not experience the problem there either, nor is it in 2.6.33.4.

Quoted 3 lines

What load figure are you basing your observations on? The 15 minutes average should be the most interesting (sampled at a 7 minutes interval...)

Quoted 12 lines

my observations are based on letting the machine idle immediately after bootup. I monitor the state of the machine using a program called conky, which I have configured to show disk I/O, cpu use, swap I/O and among other things, the load average. Immediately after booting my loadaverage tends to peak at about 2.5 to 3.0; on a working kernel this eventually settles down to 0.00 to 0.05 in about ten minutes. On kernels that exhibit this problem, it doesn't settle lower than 0.3 and is much more likely to hang anywhere from 0.8 to 1.2. In fact, if I give it enough time it'll raise and lower itself constantly without any (visible) work being done. So basically I boot the machine and go get a drink, come back, and if it's been ten minutes, there's been no disk IO, cpu use, or any other activity recorded and it's still above 0.3 something's not working right.

Do you know what load average conky is showing you? If I type 'uptime' on a console, i get three load numbers: 1minute-, 5minutes- and 15minutes-average.
If there is a systematic bias it should be visible on the 15minutes-average. If there are only bursts of 'load' it should be visible on the 1 minutes average numbers.

But it doesn't really matter for now what kind of load disturbance you are seeing, because you actually have a better way to distinguish a good kernel from a bad:

Quoted 8 lines · On Mon, 27 Sep 2010 12:32:08 -0400 [email protected] wrote:

*Something* is wrong beyond the
mere loadaverage numbers going crazy however, since timed runs of kernel compiles done with my distro's kernel and 2.6.35.5 show that while there is no *apparent* use of cpu or disk showing in vmstat while the machine is idle, the compiles on the newer kernel are taking approximately twice as long as before.

Quoted 6 lines

If you're talking about the graph,
I merely posted it to show that I've been having this problem for over a month, and it's demonstrably causing very inconsistent load averages. (Which is why the graph isn't anything close to a line, it's a mess!) the graph takes a reading every five minutes, if you were wondering about the sample rate.

Yes, the sample rate was one of the things I wanted to know, but also which of the 3 load figures you were graphing.

Quoted 3 lines

In other news, I'm in the process of bisection but keep having to skip bisects that have compile errors. sigh. still at 12 hops, somewhere around five thousands commits to check.

Good.

Quoted 3 lines

Tim McGrath

Regards,
Flo

#14 lore raw
Quoted 6 lines · Florian Mickler

Do you know what load average conky is showing you? If I type 'uptime' on a console, i get three load numbers: 1minute-, 5minutes- and 15minutes-average.
If there is a systematic bias it should be visible on the 15minutes-average. If there are only bursts of 'load' it should be visible on the 1 minutes average numbers.

It is giving the same averages that uptime does in the same format, and there is a routine problem - it remains high on all averages on the kernels that do not work properly, and zeroes eventually if I leave it alone long enough on kernels that do work properly. When I discovered X was part of the problem somehow, it was due to me testing in X with mrxvt running bash and uptime, and in console without X using bash with uptime. uptime consistently gives the same numbers that conky does, so I don't think I need to worry about conky confusing the issue.

Quoted 4 lines

But it doesn't really matter for now what kind of load disturbance you are seeing, because you actually have a better way to distinguish a good kernel from a bad:

You may think a timed kernel compile is a better way to determine if there is a fault with the kernel, but it takes my machine around two hours (WITH ccache) to build my kernel. Since the use of ccache speeds up the builds dramatically and would give misleading readings if I compiled the exact kernel source twice, I'd have to disable it if I wanted it to be a worthwhile test. So it would take even *longer* to build than normally. This is not something I'm willing to use as a 'better' test - especially since the loadavg numbers are consistently high when on a bad kernel and consistently zeroed or very close to it when not.

Here's an uptime sample from a working version:

06:20:31 up 21 min, 4 users, load average: 0.00, 0.02, 0.06

I've been typing up this email while waiting for the load to flatten from the initial boot. I think it's pretty obvious here that it's working properly, so I'm going to git bisect good it...

Bisecting: 27 revisions left to test after this (roughly 5 steps)

I'm getting fairly close at least.

Here's an uptime output from a version of the kernel that was NOT working properly, 2.6.35.6:
14:30:12 up 3:46, 4 users, load average: 0.85, 0.93, 0.89

And it probably doesn't give you any useful information, but here's 2.6.35.1's reaction to building 2.6.35:
22:01:22 up 15 min, 4 users, load average: 1.84, 1.38, 0.83

whereas on a working kernel this is what the load average looks like when building a kernel:
06:33:13 up 34 min, 4 users, load average: 1.01, 0.92, 0.52

This is not a multiprocessor or multicore system, it's an athlon XP 2800 with 1.5GB of ram. Before the question is asked, no, I'm not being silly and using make -j2.

I think simply letting the machine idle is just as good a test for determining wether or not any particular kernel is good/bad since the readings are like night and day. I only brought up that the timed kernel runs were taking longer on the kernel with the higher load average since it meant that it wasn't simply a broken statistic giving false readings; something *is* wrong, and I can't simply ignore it.

It's taken me several days to bisect this far. If greg insists, I'll restart the bisection from scratch using a kernel compile as a test, but I implore you not to ask me to do so; it will more than likely give me the same results I'm getting now for more than double the amount of time invested.

Quoted 2 lines

Yes, the sample rate was one of the things I wanted to know, but also which of the 3 load figures you were graphing.

To be honest, I actually don't know. I'm *terrible* at regex, this is what the bash script is doing:

cat /proc/loadavg | perl -p -e 's/^([^ ]+) .+$/$1/'

If you can explain what that's doing, I'd appreciate it. If it's not to your liking, I can change it to something else.

Tim McGrath

#15 lore raw
Quoted 74 lines · On Wed, Sep 29, 2010 at 1:02 PM, <[email protected]> wrote:
Quoted 6 lines · Florian Mickler

Do you know what load average conky is showing you? If I type 'uptime' on a console, i get three load numbers: 1minute-, 5minutes- and 15minutes-average.
If there is a systematic bias it should be visible on the 15minutes-average.  If there are only bursts of 'load' it should be visible on the 1 minutes average numbers.

It is giving the same averages that uptime does in the same format, and there is a routine problem - it remains high on all averages on the kernels that do not work properly, and zeroes eventually if I leave it alone long enough on kernels that do work properly. When I discovered X was part of the problem somehow, it was due to me testing in X with mrxvt running bash and uptime, and in console without X using bash with uptime. uptime consistently gives the same numbers that conky does, so I don't think I need to worry about conky confusing the issue.

Quoted 4 lines

But it doesn't really matter for now what kind of load disturbance you are seeing, because you actually have a better way to distinguish a good kernel from a bad:

You may think a timed kernel compile is a better way to determine if there is a fault with the kernel, but it takes my machine around two hours (WITH ccache) to build my kernel. Since the use of ccache speeds up the builds dramatically and would give misleading readings if I compiled the exact kernel source twice, I'd have to disable it if I wanted it to be a worthwhile test. So it would take even *longer* to build than normally. This is not something I'm willing to use as a 'better' test - especially since the loadavg numbers are consistently high when on a bad kernel and consistently zeroed or very close to it when not.

Here's an uptime sample from a working version:

06:20:31 up 21 min,  4 users,  load average: 0.00, 0.02, 0.06

I've been typing up this email while waiting for the load to flatten from the initial boot. I think it's pretty obvious here that it's working properly, so I'm going to git bisect good it...

Bisecting: 27 revisions left to test after this (roughly 5 steps)

I'm getting fairly close at least.

Here's an uptime output from a version of the kernel that was NOT working properly, 2.6.35.6:
14:30:12 up  3:46,  4 users,  load average: 0.85, 0.93, 0.89

And it probably doesn't give you any useful information, but here's 2.6.35.1's reaction to building 2.6.35:
22:01:22 up 15 min, 4 users, load average: 1.84, 1.38, 0.83

whereas on a working kernel this is what the load average looks like when building a kernel:
06:33:13 up 34 min,  4 users,  load average: 1.01, 0.92, 0.52

This is not a multiprocessor or multicore system, it's an athlon XP 2800 with 1.5GB of ram. Before the question is asked, no, I'm not being silly and using make -j2.

I think simply letting the machine idle is just as good a test for determining wether or not any particular kernel is good/bad since the readings are like night and day. I only brought up that the timed kernel runs were taking longer on the kernel with the higher load average since it meant that it wasn't simply a broken statistic giving false readings; something *is* wrong, and I can't simply ignore it.

It's taken me several days to bisect this far. If greg insists, I'll restart the bisection from scratch using a kernel compile as a test, but I implore you not to ask me to do so; it will more than likely give me the same results I'm getting now for more than double the amount of time invested.

Quoted 2 lines

Yes, the sample rate was one of the things I wanted to know, but also which of the 3 load figures you were graphing.

Quoted 7 lines

To be honest, I actually don't know. I'm *terrible* at regex, this is what the bash script is doing:

cat /proc/loadavg     | perl -p -e 's/^([^ ]+) .+$/$1/'

If you can explain what that's doing, I'd appreciate it. If it's not to your liking, I can change it to something else.

You are taking the average over the last minute.

Quoted 3 lines

Tim McGrath

#16 lore raw
In reply to [email protected]
Quoted 8 lines · On Wed, 29 Sep 2010 07:02:48 -0400 [email protected] wrote:
Quoted 6 lines · Florian Mickler

Do you know what load average conky is showing you? If I type 'uptime' on a console, i get three load numbers: 1minute-, 5minutes- and 15minutes-average.
If there is a systematic bias it should be visible on the 15minutes-average. If there are only bursts of 'load' it should be visible on the 1 minutes average numbers.

Quoted 8 lines

It is giving the same averages that uptime does in the same format, and there is a routine problem - it remains high on all averages on the kernels that do not work properly, and zeroes eventually if I leave it alone long enough on kernels that do work properly. When I discovered X was part of the problem somehow, it was due to me testing in X with mrxvt running bash and uptime, and in console without X using bash with uptime. uptime consistently gives the same numbers that conky does, so I don't think I need to worry about conky confusing the issue.

Ok. I just asked out of curiosity and to establish a common baseline.

Quoted 6 lines
Quoted 5 lines

But it doesn't really matter for now what kind of load disturbance you are seeing, because you actually have a better way to distinguish a good kernel from a bad:

Quoted 9 lines

You may think a timed kernel compile is a better way to determine if there is a fault with the kernel, but it takes my machine around two hours (WITH ccache) to build my kernel. Since the use of ccache speeds up the builds dramatically and would give misleading readings if I compiled the exact kernel source twice, I'd have to disable it if I wanted it to be a worthwhile test. So it would take even *longer* to build than normally. This is not something I'm willing to use as a 'better' test - especially since the loadavg numbers are consistently high when on a bad kernel and consistently zeroed or very close to it when not.

I didn't know that. The right thing to do is to bisect using the criteria that is easiest for you. After all the likelihood of those two symptoms (kernel build taking longer and higher load average) beeing correlated is quite high.

I termed it 'better way' because a performance regression is actually a serious issue, while load averages going crazy may or may not be serious. Also I didn't know that the difference between the load-average figures would be easy to reproduce.

In all cases, you can simply do the bisection based on the 'load average' criteria and then later check if the changeset that you've found that way also influences the kernel compile times.

Quoted 11 lines

Here's an uptime sample from a working version:

06:20:31 up 21 min, 4 users, load average: 0.00, 0.02, 0.06

I've been typing up this email while waiting for the load to flatten from the initial boot. I think it's pretty obvious here that it's working properly, so I'm going to git bisect good it...

Bisecting: 27 revisions left to test after this (roughly 5 steps)

I'm getting fairly close at least.

Out of curiosity, what region are you circling in?

Quoted 16 lines

Here's an uptime output from a version of the kernel that was NOT working properly, 2.6.35.6:
14:30:12 up 3:46, 4 users, load average: 0.85, 0.93, 0.89

I think simply letting the machine idle is just as good a test for determining wether or not any particular kernel is good/bad since the readings are like night and day. I only brought up that the timed kernel runs were taking longer on the kernel with the higher load average since it meant that it wasn't simply a broken statistic giving false readings; something *is* wrong, and I can't simply ignore it.

It's taken me several days to bisect this far. If greg insists, I'll restart the bisection from scratch using a kernel compile as a test, but I implore you not to ask me to do so; it will more than likely give me the same results I'm getting now for more than double the amount of time invested.

Quoted 4 lines
Quoted 2 lines

Yes, the sample rate was one of the things I wanted to know, but also which of the 3 load figures you were graphing.

To be honest, I actually don't know. I'm *terrible* at regex, this is what the bash script is doing:

It's taking the first column of the line in /proc/loadavg which corresponds to the 15 min avg on my machine.

Quoted 1 line

cat /proc/loadavg | perl -p -e

's/'

s before the / means, we are replacing something. After the / begins the pattern we are searching.

then the next '^' is the beginning of line marker

	'([^ ]+)'

This matches one or more non-whitespace characters. Inside a character group [] the ^ negates the character group, so that it matches everything not in that group. The plus means "one or more" and the brackets () make the substring matched by the enclosed pattern available as $1.

	' .+$/' 

The whitespace matches a whitespace, the . is a metacharacter meaning "any char" and the + specifying the quantity "one or more" again. Last the $ means 'end of line'.

So this part matches any whitespaced followed by at least one character.

In effect the whole expression replaces everything on the line with the first consecutive non-blank character sequence, iff there are non-blank characters at the beginning of the line.

One could also just use awk for this. ( awk '{print $1}' /proc/loadavg)

Quoted 3 lines

If you can explain what that's doing, I'd appreciate it. If it's not to your liking, I can change it to something else.

No, it's ok.

Regards,
Flo

#17 lore raw
Quoted 10 lines · On Wed, 29 Sep 2010 13:52:48 +0200 Florian Mickler <[email protected]> wrote:
Quoted 4 lines · On Wed, 29 Sep 2010 07:02:48 -0400 [email protected] wrote:
Quoted 2 lines

Yes, the sample rate was one of the things I wanted to know, but also which of the 3 load figures you were graphing.

To be honest, I actually don't know. I'm *terrible* at regex, this is what the bash script is doing:

It's taking the first column of the line in /proc/loadavg which corresponds to the 15 min avg on my machine.

Actually, no. It's the 1 min avg...

Try using

awk '{print $3}' /proc/loadavg

but I don't think it necesserily invalidates your bisection results.

Regards,
Flo

#18 lore raw
In reply to Florian Mickler
Quoted 3 lines · Florian Mickler

In all cases, you can simply do the bisection based on the 'load average' criteria and then later check if the changeset that you've found that way also influences the kernel compile times.

Works for me.

Quoted 1 line

Out of curiosity, what region are you circling in?

git log says I'm at
commit e7858f52a5cb868289a72264534a1f05f3340c6c
Merge: 27a9da6 bbf1bb3
Author: Ingo Molnar <[email protected]>
Date: Sat May 8 18:11:19 2010 +0200

Merge branch 'cpu_stop' of
git://git.kernel.org/pub/scm/linux/kernel/git/tj/misc into sched/core

but I don't know if this is the right command. I'm almost entirely a novice at git... Anyway, I haven't tested if this particular commit works yet, I'll be rebooting after I send this email.

Quoted 1 line

One could also just use awk for this. ( awk '{print $1}' /proc/loadavg)

Thanks for explaining what it's doing. I assume if I wanted to use the third argument I'd use

	awk '{print $3}' /proc/loadavg

correct? I think I'll be using that since it'll give me a better clue if loads are consistently high over a long period, as well as make the loadavg graph less prone to jitter.

Thank you,
Tim McGrath

#19 lore raw
Quoted 8 lines · On Wed, 29 Sep 2010 21:33:05 -0400 [email protected] wrote:
Quoted 3 lines · Florian Mickler

In all cases, you can simply do the bisection based on the 'load average' criteria and then later check if the changeset that you've found that way also influences the kernel compile times.

Works for me.

Quoted 1 line

Out of curiosity, what region are you circling in?

git log says I'm at

git-log is not that expressive here, as it only shows the history of current HEAD (the currently checked out version). You can post the output of 'git bisect log'.
If you wanna know how the region looks like you can type 'git bisect visualize'.

Regards,
Flo

#20 lore raw
Quoted 5 lines · Florian Mickler

git-log is not that expressive here, as it only shows the history of current HEAD (the currently checked out version). You can post the output of 'git bisect log'.
If you wanna know how the region looks like you can type 'git bisect visualize'.

Wow that's neat. I can't be sure I understand exactly what its saying but if I'm looking at it right I'm in the middle of a temporary fork that was created just after 2.6.34-rc6 and was merged with 2.6.35-rc1.

This is what my git bisect log looks like:

tm@roll:~/src/linux-2.6$ git bisect log
git bisect start
# good: [e40152ee1e1c7a63f4777791863215e3faa37a86] Linus 2.6.34
git bisect good e40152ee1e1c7a63f4777791863215e3faa37a86
# bad: [9fe6206f400646a2322096b56c59891d530e8d51] Linux 2.6.35
git bisect bad 9fe6206f400646a2322096b56c59891d530e8d51
# skip: [e0bc5d4a54938eedcde14005210e6c08aa9727e4] Merge branch 'i2c-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging
git bisect skip e0bc5d4a54938eedcde14005210e6c08aa9727e4
# good: [c72ea83d39efa71bf0e3fc1125e780071a79f1bc] e1000e: Remove unnessary log message
git bisect good c72ea83d39efa71bf0e3fc1125e780071a79f1bc
# skip: [4403158ba295c8e36f6736b1bb12d0f7e1923dac] Ban ecryptfs over ecryptfs
git bisect skip 4403158ba295c8e36f6736b1bb12d0f7e1923dac
# bad: [c392c4c6dccf7c64c113b473c7eceedf25eddd51] Merge branch 'perf' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 into perf/core
git bisect bad c392c4c6dccf7c64c113b473c7eceedf25eddd51
# good: [f869097e884d8cb65b2bb7831ca57b7dffb66fdd] perf session: Make read_build_id routines look at the host_machine too
git bisect good f869097e884d8cb65b2bb7831ca57b7dffb66fdd
# bad: [b8ae30ee26d379db436b0b8c8c3ff1b52f69e5d1] Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
git bisect bad b8ae30ee26d379db436b0b8c8c3ff1b52f69e5d1
# good: [1014cfe2fb4cdd663137aafb21448cb613dd6a7d] Merge branch 'core-locking-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
git bisect good 1014cfe2fb4cdd663137aafb21448cb613dd6a7d
# good: [3aaf51ace5975050ab43c7d4d7e439e0ae7d13d7] Merge branch 'oprofile-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
git bisect good 3aaf51ace5975050ab43c7d4d7e439e0ae7d13d7
# good: [371fd7e7a56a5c136d31aa980011bd2f131c3ef5] sched: Add enqueue/dequeue flags
git bisect good 371fd7e7a56a5c136d31aa980011bd2f131c3ef5
# bad: [e7858f52a5cb868289a72264534a1f05f3340c6c] Merge branch 'cpu_stop' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/misc into sched/core
git bisect bad e7858f52a5cb868289a72264534a1f05f3340c6c

Just a few more steps until we find the culprit, it said 7 revisions remained after this bisect.

Tim McGrath