Linux 2.6.10-rc3 has been released
ChangeLog:
Summary of changes from v2.6.10-rc2 to v2.6.10-rc3 ============================================ [WATCHDOG] i8xx_tco.c-request_region-patch Fix: in i8xx_tco.c, during the initialisation, the driver accesses io without checking if the port is free. [NETFILTER]: Fix multiple bugs in ipv6header match This patch fixes following bugs in ip6t_ipv6header.c - The cast of the pointer to the next IPv6 extension header is wrong. - The logical operation is wrong. These fixes intends - soft mode without invert flag "!" match if the packet contains all of the specified headers. - soft mode with invert flag "!" match if the packet DOESN'T contain all of the specified headers. - strict mode without invert flag "!" match if the packet contains JUST ONLY the specified headers. if the packet doesn't contain some specified headers or contains unspecified headers, the packet doesn't match with rule. - strict mode with invert flag "!" NOT MATCH if the packet contains JUST ONLY the specified headers. Otherwise, match. So, if the packet contains some specified headers and DOESN'T contain other specified headers, the packet MATCHES with rule. Signed-off-by: Yasuyuki KOZAKAI Signed-off-by: Patrick McHardy [NETFILTER]: introduce skb_header_pointer() to ipv6header match Signed-off-by: Yasuyuki KOZAKAI Signed-off-by: Patrick McHardy [NETFILTER]: Make eui64 match usuable in FORWARD chain Signed-off-by: Yasuyuki KOZAKAI Signed-off-by: Patrick McHardy [IA64] Allocate syscall #1270 for waitid syscall. Signed-off-by: Tony Luck [ACPI] acpi_pci_irq_enable() now returns 0 on success. This bubbles all the way up to pci_enable_device(). This allows IRQ0 to be used as a legal PCI device IRQ. The ES7000 uses an interrupt source override to assign pin20 to IRQ0. Then platform_rename_gsi assigns pin0 a high-numbered IRQ -- available for PCI devices. But IRQ0 needs to be a legal PCI IRQ in the lookup code to make it as far as the re-name code. Signed-off-by: Natalie Protasevich Signed-off-by: Len Brown fix non-ACPI IOAPIC build [SPARC]: In cg3 driver, access control reg using byte not long IOs. Signed-off-by: David S. Miller [SPARC64]: Update defconfig. Signed-off-by: David S. Miller [SUNZILOG]: Update timeout when setting termios. Signed-off-by: David S. Miller Merge [IA64] speedup ptrace by avoiding kernel-stack walk This patch changes the syscall entry path to store the current-frame-mask (CFM) in pt_regs->cr_ifs. This just takes one extra instruction (a "dep" to clear the bits other than 0-37) and is free in terms of cycles. The advantage of doing this is that it lets ptrace() avoid having to walk the stack to determine the end of the user-level backing-store of a process which is in the middle of a system-call. Since this is what strace does all the time, this speeds up strace quite a bit (by ~50%). More importantly, it makes the syscall vs. non-syscall case much more symmetric, which is always something I wanted. Note that the change to ivt.S looks big but this is just a rippling effect of instruction-scheduling to keep syscall latency the same. All that's really going on there is that instead of storing 0 into cr_ifs member we store the low 38 bits of ar.pfs. Signed-off-by: David Mosberger Signed-off-by: Tony Luck [ACPI] handle out of spec EC bit width http://bugzilla.kernel.org/show_bug.cgi?id=1744 Signed-off-by: Luming Yu Signed-off-by: Len Brown [ACPI] migrate to seq_file() interface http://bugzilla.kernel.org/show_bug.cgi?id=3333 Signed-off-by: David Shaohua Li Signed-off-by: Len Brown [ACPI] automatic workaround for NFORCE timer-override BIOS bug http://bugzilla.kernel.org/show_bug.cgi?id=3551 Signed-off-by: Andi Kleen Signed-off-by: Len Brown [ACPI] clean up the NFORCE BIOS bug workaround delete now obsolete dmi_scan entries fix build for ACPI & !IOAPIC Signed-off-by: Len Brown [IA64] add cpu_relax() in the body of spin loops This patch adds cpu_relax() in the body of spin loops in smp_call_function(), smp_call_function_single(), and ia64_mca_wakeup_ipi_wait(). Signed-off-by: Fenghua Yu Signed-off-by: Tony Luck [ACPI] fix warnings for 64-bit video build (Luming Yu) Cset exclude: torvalds@ppc970.osdl.org|ChangeSet|20041111002817|28673 Cset exclude: acme@conectiva.com.br[torvalds]|ChangeSet|20041111002501|29509 [SPARC]: Remove unnecessary pm_idle comment. Signed-off-by: David S. Miller [CG14]: Fix NULL sbus_dev handling and colormap setup. Signed-off-by: David S. Miller [ACPI] remove acpi_ksyms.c Signed-off-by: Adrian Bunk Signed-off-by: Len Brown [ACPI] remove acpi_ksyms.c Signed-off-by: Adrian Bunk Signed-off-by: Len Brown [ACPI] add #ifdef ACPI_FUTURE_USAGE Signed-off-by: Adrian Bunk Signed-off-by: Len Brown [NETFILTER]: Handle nonlinear skbs in ip_queue/ip6_queue Signed-off-by: Patrick McHardy [NET]: Move rx timestamp functions to net/core/dev.c Signed-off-by: Patrick McHardy [NETFILTER]: Enable rx timestamps in ip_queue/ip6_queue Signed-off-by: Patrick McHardy [NETFILTER]: associate locally generated icmp errors with conntrack of original packet Signed-off-by: Patrick McHardy [NETFILTER]: Fix invalid tcp/udp checksums within NATed icmp errors Signed-off-by: Patrick McHardy [NET] Assign inet transport sockets to the right module This is another case where when all the protocols use sk->sk_prot we can remove another member from struct sock, sk->sk_owner, instead using the sk->sk_prot->owner, i.e. this property is not per socket instance, but per protocol. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller [PATCH] alpha sysrq compile fix missing extern declaration in -Werror land... Add the proper declaration to sysrq.h. [PATCH] sparse: add -m64 to CHECKFLAGS on alpha and sparc64 Get the sizes right, and thus avoid complaints about big constants etc. [PATCH] revert recent futex_wait fix The patch was wrong. Back it out, and add some commentary explaining why we need to run queue_me() prior to the get_user(). Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [libata] fix DocBook bugs [ARM PATCH] 2234/3: S3C2410 - new serial driver (1/4) Patch from Ben Dooks S3C2410/S3C2440 new serial driver this patch includes the following - clock selection using clock core - handling for both 2410 and 2440 uarts - uart suspend/resume support - Dimitry Andric's fix for un-initialised spinlocks - Herbert Poetzl's fixes for the following -> break character recognition -> Magic-SYSRQ handling -> Uart software flow control (IrDA console) Updated to include rmk's comments from Patch #2234/1 and fix a few checks for serial clocks, as well as merging fixes from rmk that had taken time to go through the biteepr to release. Signed-off-by: Ben Dooks Signed-off-by: Russell King [SERIAL] s3c2410: remove duplicate include [ARM PATCH] 2241/1: S3C2410 - default configuration update Patch from Ben Dooks updates to the default configuration to change the following: - use S3C2410 internal rtc instead of PC style rtc - add mtd support for nand - add mtd support for nor (and bast nor mapping) Signed-off-by: Ben Dooks Signed-off-by: Russell King [ARM PATCH] 2242/1: BAST - default configuration update Patch from Ben Dooks Updates for the Simtec BAST (EB2410ITX) default configuration to fix the following: - add mtd support for NAND and NOR - remove PC style parallel port (does not build) - use S3C2410 core RTC for time - remove non-bast S3C2410 machines Signed-off-by: Ben Dooks Signed-off-by: Russell King [ARM PATCH] 2243/1: BAST - move pm init to init_machine Patch from Ben Dooks Move the pm code to use the machine's init_machine code, instead of an late_initcall(). This change is as-per the documentation, and means we do not need to check for machine type in the init code. Thanks to Dimitry Andric for pointing this out Signed-off-by: Ben Dooks Signed-off-by: Russell King [ARM PATCH] 2246/1: S3C2410 - rename i2c depending on 2410/2440 Patch from Ben Dooks Ensure the i2c platform device has the correct name if the system is an s3c2440. Signed-off-by: Ben Dooks Signed-off-by: Russell King [ARM PATCH] 2247/1: S3C2410 - serial low-level updates Patch from Ben Dooks Move the configuration for the UART to use for the low-level messages generated by the uncompressor out of the kernel debug section, which means the system builds properly without having kernel debug enabled. Also remove the use of including headers, and the default configuration updates Signed-off-by: Ben Dooks Signed-off-by: Russell King [PATCH] Regparm for x86 machine check handlers The patch to change traps and interrupts to the fastcall convention missed the machine check handlers. Signed-off-by: Brian Gerst Signed-off-by: Linus Torvalds [NETFILTER]: revert MASQUERADE optimization for mostly static IPs Signed-off-by: Phil Oester Signed-off-by: Patrick McHardy [AF_UNIX]: Serialize dgram read using semaphore just like stream. Signed-off-by: David S. Miller [IPV6]: Temp fix for ipv6 link-local address problem. Make sure loopback_dev, if up, has the ipv6 bits for it setup before the addrconf netdev notifier is registered. Signed-off-by: David S. Miller [NETLINK]: Fix mc_list operations While doing the 2.4 version of the netlink patch, I noticed a bug in my mc_list code. First of all it wasn't holding the table lock when adding nodes to the list. It also didn't take the node off the list when multicast is switched off. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller [NETFILTER]: Cleanup find_appropriate_src() Fix The fix to find_appropriate_src left a stupid test, which has no effect but is incoherent at best. It covers a corner case: if a previous connection from this source was mapped onto a different IP address (because it was explicitly told to), should the next one be mapped that was as well, if no explicit rule says to remap the src IP? I think the answer is yes: unless the user explicitly tells us to map into a particular range, we should follow the Kegel draft. Signed-off-by: Rusty Russell Signed-off-by: David S. Miller [NETFILTER]: Fix stack leakage in iptables Found running netfilter code under valgrind: we return some uninitialized stack to userspace (needs root). Almost certainly harmless, but this suppressed the valgrind complaint, and doesn't hurt. Signed-off-by: Rusty Russell Signed-off-by: David S. Miller [SCTP]: Fix inetaddr notifier chain corruption Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller [PATCH] fix appletalk locking Just tried the new toy. It works. [TG3]: 5753 support and a bug fix. Add support for 5753 chips which is mostly just adding in the appropriate PCI ids and recognizing that these chips do not use GPIO2 for Vaux switching. Also do not set DMA read water mark on PCI Express. In such configurations these bits are reserved. Signed-off-by: David S. Miller [TG3]: Update driver version and reldate. Signed-off-by: David S. Miller [IPV6]: Fix xfrm6_tunnel_check_size mtu calc. OK, let's make it the same as IPv4. This is still broken as the calculation is wrong. However, to fix it properly we need to store the MTUs inside the dst's. I'll get around to that sooner or later. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller [PATCH] x86-64: Fix get_user_pages access to vsyscall page The current kernel oopses on x86-64 when gdb steps into the vsyscall page. This patch fixes it. I also removed the bogus NULL checks of *_offset and replaced them with proper _none checks. I made them BUGs because vsyscall pages should be always mapped. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] mm: tune the page allocator thresholds without patch: pages_min pages_low pages_high dma 4 8 12 normal 234 468 702 high 128 256 384 with patch: pages_min pages_low pages_high dma 17 21 25 normal 939 1173 1408 high 128 160 192 without patch: | GFP_KERNEL | GFP_ATOMIC allocate immediately | 9 dma, 469 norm | 9 dma, 469 norm allocate after waking kswapd | 5 dma, 234 norm | 3 dma, 88 norm allocate after synch reclaim | 5 dma, 234 norm | n/a with patch: | GFP_KERNEL | GFP_ATOMIC allocate immediately | 22 dma, 1174 norm | 22 dma, 1174 norm allocate after waking kswapd | 18 dma, 940 norm | 6 dma, 440 norm allocate after synch reclaim | 18 dma, 940 norm | n/a So the buffer between GFP_KERNEL and GFP_ATOMIC allocations is: 2.6.8 | 465 dma, 117 norm, 582 tot = 2328K 2.6.10-rc | 2 dma, 146 norm, 148 tot = 592K patch | 12 dma, 500 norm, 512 tot = 2048K Which is getting pretty good. kswap starts at: 2.6.8 477 dma, 496 norm, 973 total 2.6.10-rc 8 dma, 468 norm, 476 total patched 17 dma, 939 norm, 956 total So in terms of total pages, that's looking similar to 2.6.8. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] low discontig highmem_start_page In the case of i386 CONFIG_DISCONTIGMEM CONFIG_HIGHMEM without highmem, highmem_start_page was wrongly initialized (from a NULL zone_mem_map), causing __change_page_attr to BUG on boot. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] tmpfs free_inodes leak When new_inode failed, shmem_get_inode forgot to restore free_inodes. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc32: Fix Motorola Sandpoint builds Motorola Sandpoint builds broke recently, as part of the pci_find_device -> pci_get_device change. The following is the trivial fix. Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: iSeries legacy model emulation of PURR Here's a patch to extend the current Linux on Power support for PURR to legacy IBM iSeries servers (pre-Power5 processor models). This patch enables the reporting of timebase metrics to reflect physical processor utilization in a system running multiple logical partitions which share the same physical processors. The patch simply uses existing user interfaces for Linux IBM Power5 based servers to report data already collected by the hypervisor. The values reported with each call are running values in units of the system timebase. The calculation of physical processor utilization results from two samples (purr1 and purr2) differing by a know interval (time) such that: physical utilization = (purr2 - purr1) / (time * number of procs * timebase) where the number of procs and timebase can be obtained from /proc/cpuinfo. Applications have been written to the interface already defined and these applications have value back on the legacy iSeries models. Signed-off by: Jeff Scheel (scheel at vnet.ibm.com) Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: Kill unused KRANGE_{START,END} macros Remove KRANGE_{START,END} macros from ppc64 code. These were not used anywhere. Further KRANGE_END was misleading, since it implied a limit on the linear mapping range based on the pagetable structure, whereas in fact the linear mapping does not use a (Linux) pagetable at all. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: ratelimit some rtas errors Use printk_ratelimit() in rtc code to avoid flooding the kernel log buffer with errors. Also use rtas_get_error_log_max() instead of duplicating it in __fetch_rtas_last_error. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: Use pci_device_to_OF_node PCI_GET_DN() doesnt check to see if ->sysdata has been initialised correctly - we should instead use pci_device_to_OF_node. Leave PCI_GET_DN() in the one performance critical case (iommu table lookup in pci DMA functions). In this case ->sysdata is guaranteed to have been initialised by the iommu setup code. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: avoid 32bit only syscalls in unistd.h Steve Munroe points out that ppc64 glibc builds stubs for a number of 32bit only syscalls. While none of them exist in the kernel syscall table, their existence in unistd.h means glibc still tries to use them then falls back onto the 64bit safe versions. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: reduce ifdef clutter in arch/ppc64/kernel/sysfs.c From: Christoph Hellwig Reduce ifdef clutter in arch/ppc64/kernel/sysfs.c Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: pci cleanup Cleanup ppc64 pci code. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: remove phb_set_model phb_set_model does a lot of work just to set up a text string that almost nothing uses. Replace this all with an is_python() check. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: make fixup_winbond_82c105 pseries specific The winbond irq fixup is pSeries specific. Move it into pSeries_pci.c and check for PLATFORM_PSERIES. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: remove duplication in pci_alloc_* We duplicated the code in pci_alloc_pci_controller twice and had an ifdef for iseries as well, just to select between kmalloc and bootmem memory. Change this so we instead pass the allocation into a common function - pci_setup_pci_controller. Also use a spinlock around the host_list and global_phb_number code since we now can modify it at runtime via hotplug. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: OF overrides for pci_probe_only, pci_assign_all_buses Allow pci_probe_only and pci_assign_all_buses to be modified via OF properties. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: remove BUG()s in pcibios_fixup_bus BUG() on missing IO or memory resources in pcibios_fixup_bus is rude, remove it. Also use list_for_each_entry. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: get_phb_reg_prop only required on python PCI machines get_phb_reg_prop was only used for python PCI machines, so remove it from common code and call it from there. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: alloc_bootmem returns void * As Andrew points out, there is no need for casts since alloc_bootmem returns void*. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: cleanups hpte_init_native, kill warning for !PSERIES builds From: Christoph Hellwig This splits out a small helper that checks whether tlb batching should be enabled from hpte_init_native, thus cleaning up the ifdef hell and killing a warning for pmac builds. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] PPC64 call ibm,os-term only if its available This patch is from Olaf Hering. The rtas property 'ibm,os-term' is not available on JS20, a panic will print: unable to mount root filesystem on /dev/hda Kernel panic - not syncing: Attempted to kill init! <0>ibm,os-term call failed -1 Rebooting in 42 seconds.. Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] PPC64 rtasd: window when error_log_cnt could get zeroed This patch is from Jake Moilanen . There appears to be a hole that if we get an log_error() call, that we could zero out our error log count in nvram. When rtasd() starts up, it turns on the logging via 'no_more_logging = 0'. If we get a log_error() call after that is set but before nvram_read_error_log has actually read nvram to set error_log_cnt, the log_error() call will write back to nvram a uninitialized error_log_cnt value, and wipe out our sequence number. To close the hole, simply move the 'no_more_logging = 0' till after nvram sets error_log_cnt but before pSeries_log_error is called. I also changed the 'no_more_logging' variable to be 'no_logging' since it's not only used when we stop logging now. I also removed the "volatile" part of no_more_logging, since it's unneeded. Signed-off-by: Jake Moilanen Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64 iSeries: don't share request queues in viocd This patch fixes the virtual cdrom driver to not share a single request queue. Sharing the queue causes an oops if you remove the module and more than one cdrom exists. Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: add missing braces to rtc driver This patch fixes the PPC64 rtc driver where a pair of braces was missing. Not a big bug, but a bug none the less. Also, while I was there, use C99 initialisers. Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] add cpu_relax() in spin loops & clean up barrier() The patch adds cpu_relax() in the body of some spin loops for 2.6.9. The patch also removes redundant barrier() code after cpu_relax() on ia32. In the PAUSE instruction section, IA32 SDM claims "it is recommended that a PASUE instruction be placed in all spin-wait loops". And x86_64 SDM says that PAUSE instruction is same as legacy mode in IA-32e mode operation. This patch is against 2.6.9 (kernel.org). It was tested on ia32 and x86_64. Acked-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] i386: apic_printk() used before initialized Both detect_init_APIC() and init_apic_mappings() it's called from are invoked early, before the command line has been processed. Therefore its meaningless to call apic_printk() from them as that depends on apic_verbosity which is initialized from the command line. I could move apic_verbosity initialization to parse_cmdline_early(), but I think that would be an overkill, the point being we are initerested in feedback from detect_init_APIC() anyway. Without that it's hard to tell what's really going on as it's been the case with the recent report of the local APIC being non-functional despite the whole setup being apparently correct. So I converted these calls to ordinary printk() invocations. The init_apic_mappings() are less interesting, so I've made them output at the debug level. While at it I've made some obvious nearby formatting clean-up. Signed-off-by: Maciej W. Rozycki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] x86_64 hpet: fix function warning put function prototype outside of #ifdef block, to fix: arch/x86_64/kernel/time.c:941: warning: implicit declaration of function `oem_force_hpet_timer' Signed-off-by: Randy Dunlap Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] H8/300: /proc/cpuinfo typo fix Signed-off-by: Yoshinori Sato Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] H8/300: signal handling update Signed-off-by: Yoshinori Sato Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] H8/300: read{b,w,l} / write{b,w,l} error fix read{b,w,l} and write{b,w,l} compile error fix. Signed-off-by: Yoshinori Sato Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] H8/300: vmlinux.lds.S update - duplicate define section delete. - fix CONFIG_ROMKERNEL. Signed-off-by: Yoshinori Sato Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: update for m32r-g00ff This patchset was originally from NIIBE Yutaka. These patch update the m32r kernel for a new bootloader "m32r-g00ff". The "m32r-g00ff" has been written and developed by NIIBE Yutaka, and released under the GPL from http://www.gniibe.org/. So far, it supports two types of booting operations, CF boot and Network boot (HTTP boot). * CF boot - boot from CompactFlash or Microdrive(TM) We can boot a kernel from CF device. To make use of m32r-g00ff, we just put a first stage IPL(initial program loader) into a flash memory, and a secondary bootloader into CF media device. Currently, LILO-21.4.4 can be used to write m32r-g00ff into the boot sector of CF device on a cross development environment. * HTTP boot - boot via network with HTTP protocol By using m32r-g00ff, we can download and boot a kernel image from a web-server. The m32r-g00ff downloads a kernel image from a given URL, resolving the webserver's IP address by DNS. As a preparation, we just place a secondary bootloader binary and a kernel image on the webserver. - Position-independent zImage support; this aims at removing constraints of zImage(vmlinuz)'s location. Signed-off-by: NIIBE Yutaka Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: CF boot support for Mappi2 - Update io_mappi2.c to access a CF device as an IDE disk device for Mappi2 eva board. - Please set CONFIG_M32R_CFC=n, when you use m32r-g00ff for CF boot. Signed-off-by: NIIBE Yutaka Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: update defconfig files Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: update dot.gdbinit files This patch is for updating GDB initalization files (dot.gdbinit*) for all m32r target boards. Currently, part of dot.gdbinit* files are maintained by using a "gen_gdbinit" script. http://www.linux-m32r.org/eng/download.html#othertools Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: signal bug fix This patch fixes a bug introduced in the last batch of signal fixes. The system call return value should only be reset if called diectly from a system call, i.e. sigsuspend. The fixes added earlier caused any interrupted non-zero system call return to be reset, confusing fork and vfork, among others. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: 64-bit cleanups in the system calls 64-bit cleanup - this fixes the return values of the system calls to be longs. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: 64-bit type cleanups 64-bit cleanliness - Fix the number of bits of the time_t field in the COW header to be 32 and change an int to a longs. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: don't rule out syscall_nr == 0 From: Bodo Stroesser Change the valid system call numbers to reflect the possibility that we could have __NR_restart_syscall. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: redundant code removal from signal delivery From: Bodo Stroesser Change the do_signal interface to eliminate its argument. Also, remove the calls from the system call handlers since they are redundant. In all cases, pending signals are checked for in the interrupt handler. Temporarily, do_signal passes the current error to kern_do_signal. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: redundant argument removal from handle_signal From: Bodo Stroesser Change the interface to handle_signal so that it doesn't take the system call return value as an argument and eliminate its return value. kern_do_signal also now doesn't return immediately after determining that there is no signal to deliver. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: handle_signal simplification From: Bodo Stroesser Move the signal delivery code around to eliminate a couple of temporary variables. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: fix setting of interrupted syscall return value From: Bodo Stroesser handle_signal now checks whether it is being called from a system call invocation. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: make signal frame construction more resemble x86 From: Bodo Stroesser This makes the UML signal frame construction interface somewhat more similar to x86 than before. Also, some small code cleanup, and checking for errors before changing the signal mask in the SA_NODEFER case. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: fix signal mask on delivery error From: Bodo Stroesser If the user stack limit is reached or the signal stack assigned with sigaltstack() is invalid when a user signal handler with SA_ONSTACK has to be started, the signal mask of the interrupted user program is modified. This happens because the mask, that should be used with the handler only, is written to "current->blocked" even if the handler could not be started. But without a handler, no rewrite of the original mask at sys_sigreturn will be done. A slightly different case is sys_sigsuspend(), where the mask is already modified when kern_do_signal() is started. "*oldset" and "current->blocked" are not equal here and thus current->blocked has to be set to *oldset, if an error occurs in handle_signal(). For both cases I've written small tests, and with the patch the result is OK. This issue is relevant for other architectures too (e.g. i386, I've seen). Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: Don't delay segfaults From: Bodo Stroesser This one covers the fact, that the SIGSEGV signal, which is created by force_sigsegv() in case of an error in handle_signal(), is not delivered to the user immediately. In the worst case it even could be masked if a sigprocmask() systemcall follows immediately after return from kernel. The patch is relevant for other architectures, too. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: fix definitions of pte_unmap_* Some definitions of pte_unmap_* macros were written for HIGHPTE, which UML doesn't support. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: LFS 64-bit cleanups Add ARCH_USER_CFLAGS so the arches can modify USER_CFLAGS. We now take __ARCH_WANT_OLD_STAT and __ARCH_WANT_STAT64 from the base arch so that the LFS-64 code gets included or excluded automatically. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: Remove unused declaration Remove an unused declaration of time_stamp(). Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: remove some dead code Bodo pointed out that arch/um/kernel/skas/exec_user.c is dead code, so this removes it. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: defconfig update Update defconfig. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: core changes From: Stefan Bader From: Martin Schwidefsky s390 core changes: - Store correct set of registers to core dumps. - Fix make install with separate obj directory. - Regenerate default configuration. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: common i/o layer From: Cornelia Huck From: Thomas Spatzier common i/o layer changes: - Prevent double unregister of ccw devices. - Move unregister out of the subchannel remove function, to avoid live-lock due to hotplug if the root device is currently indisposed. - Delete pending timer after a machine check. - Revert change to allocate qdio queues and SLIBS in DMA memory. - Decrement ccw_device_init_count only after ccw_device_register is done. - Remove unnecessary check in ccw_hotplug. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: dasd driver From: Stefan Weinhuber dasd driver changes: - Fix parameter parsing to allow sequences of devices, ranges and keywords. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: 3270 console 3270 console driver changes: - Add error handling in 3270 device startup. - Do halt_io if startup has been interrupted. - Fix reference counting in tty timers. - Simplify set_timer functions. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: monreader docu From: Gerald Schaefer Docu for the z/VM monitor record read access feature. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: crypto driver From: Eric Rossman s390 crypto driver changes: - Small cleanup: misc -> crypto, header file defines, variable names and a printk message. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: zfcp host adapter From: Andreas Herrmann zfcp host adapter change: - Avoid usage of unregister debug feature. - Avoid race when unregistering debug feature. - Corrected some log messages for WKA ports. - Don't pass NULL pointer to debug_register_view and debug_set_level. - Some coding style cleanup. - Fix race between scsi_add_device and deregistration of the adapter. - Shorten & rename zfcp_els/zfcp_els_handler. - Remove unused code for unused ELS commands. - Evaluate response instead of request in adisc handler. - Allocate qdio queue structures below 2GB. - Remove ifdefs around ioctl32.h. - Use CONFIG_COMPAT instead of CONFIG_S390_SUPPORT. - Use semaphore in zfcp_ccw_shutdown. - Strip down debug_register/debug_unregister. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: zfcp read-only lun sharing From: Volker Sameske zfcp host adapter: - Add read-only lun sharing feature. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: zfcp act enhancements From: Andreas Herrmann From: Maxim Shchetynin zfcp host adapter changes: - Add access control enhancements. - Add event callbacks. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] radix_tree_delete() fix I was looking through the radix tree code and came across what I think is a bug in radix_tree_delete. for (idx = 0; idx < RADIX_TREE_TAG_LONGS; idx++) { if (pathp[0].node->tags[tag][idx]) { tags[tag] = 1; nr_cleared_tags--; break; } } The above loop should only be executed if tags[tag] is zero. Otherwise, when walking up the tree, we can decrement nr_cleared_tags twice or more for the same value of tag, thus potentially exiting the outer loop too early. Ensure that nr_cleared_tags is only decremented once for each tag. Signed-off-by: Dave Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] kNFSd: fix d_find_alias brokenness 10 weeks ago, http://linux.bkbits.com:8080/linux-2.5/cset@415b3380pxf4sB97gM8ujLqDxi6G... The patch was mostly right, and fixed a real problem, but missed a bit. It passed the job of checking if an inode had a current alias off to d_find_alias instead of open-coding it in d_alloc_anon. However there is one case where d_alloc_anon would not return the right dentry. That case being when the inode was for the root of the filesystem. The root is a special case because it is not hashed. All other dentries that are not hashed are quite un-interesting: There are "unlinked" but not yet closed. The root of a filesystem is unhashed, but is interesting. Allowing d_find_alias to return an unhashed alias for a directory solves this problem. It is safe because callers of d_find_alias on a directory inode either have a name for the inode already (so finding an unlinked directory by mistake is impossible) or will soon be looking for a name and will drop the dentry if a name is not found. Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] kill lockd_syms.c The patch below kills lockd_syms.c. Signed-off-by: Adrian Bunk Acked-by: Trond Myklebust Acked-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Don't remove /sys in initramfs Using the "resume" kernel parameter together with an initramfs revealed a bug that causes removal of the /sys directory in the initramfs' tmpfs, making the system unbootable. The source of the problem is that the try_name() function removes the /sys directory unconditionally, instead of removing it only when it has been created by try_name(). The attached patch only removes /sys if it has been created before. Signed-off-by: Juerg Billeter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] sound_alloc_dmap memory allocation warning suppression alloc_pages() failures are expected here. Manually prevent warnings. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] CONFIG_UNIX98_PTY=n warning fix ptmx_open() only exists if CONFIG_UNIX98_PTYS. Signed-off-by: Yoshinori Sato Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Fix pmac_zilog.c so it compiles again It seems that pmac_zilog.c got missed in the dev.power_state to dev.power.power_state conversion. This patch makes that change, and also fixes a problem where it would not compile if CONFIG_MAGIC_SYSRQ was set but CONFIG_SERIAL_CORE_CONSOLE was not. Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Do power_state conversion for mesh.c This patch changes dev.power_state to dev.power.power_state in drivers/scsi/mesh.c, and fixes an uninitialized variable use in a printk. Signed-off-by: Paul Mackerras Cc: James Bottomley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Add __iomem annotations to drivers/scsi/mac53c94.c This patch adds __iomem annotations to drivers/scsi/mac53c94.c, and changes one use of st_le32 to writel. Signed-off-by: Paul Mackerras Cc: James Bottomley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] power_state and __iomem for mediabay.c This patch does the power_state -> power.power_state conversion for drivers/macintosh/mediabay.c and makes it use void __iomem * for ioremap cookies. Once the IDE code is converted to not use unsigned long for MMIO register addresses, I will be able to remove a few casts from here. Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] __iomem annotations for swim3.c This patch adds __iomem annotations to drivers/block/swim3.c. Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Multilink fix for ppp_generic.c I released ppp-2.4.3 yesterday, with much improved support for multilink operation - when the first link is terminated, its pppd no longer exits immediately, but keeps running in order to keep the ppp interface up while there are other links still in the bundle. However, this shows up a bug in the kernel ppp driver, which is that there is no way for the pppd controlling the bundle to know when the last link in the bundle is terminated. This patch provides such a way: with this patch, pppd will get an EOF when reading from the /dev/ppp instance for the bundle when there are no channels connected. The change does not affect older versions of pppd or normal non-multilink operation (I have tested to make sure of that). Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] cdev_init: zero out cdev before kobject_init() Right now, cdev_init() works in a way that is not very intuitive. If a driver passes an uninitialized struct cdev to cdev_init(), then an uninitialized struct kobject will be passed to kobject_init(), which does kset_get() on kobj->kset, which probably points off into space and causes an oops. Drivers can work around this by zeroing out their struct cdev in advance (and indeed most if not all of the things passed to cdev_init() come from BSS) but I think it makes more sense for cdev_init() to live up to its name and actually work on an uninitialized cdev. Signed-off-by: Roland Dreier Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] dm_init unresolved reference to _exits drivers/md/dm.c dm_int refers to _exits which is defined as __exitdata. With CONFIG_HOTPLUG=n, __exitdata is discarded. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] device-mapper: dm-crypt fix for zero-length key dm-crypt fix for zero-length key. Signed-Off-By: Christophe Saout Signed-Off-By: Alasdair G Kergon Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] device-mapper: Fix some DMERR macro usage Fix some DMERR macro usage. It already adds : and \n. Signed-Off-By: Alasdair G Kergon Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] device-mapper: Add DM_TARGET_MSG Add DM_TARGET_MSG ioctl so data can be passed to a dm target after its table has been loaded. Signed-Off-By: Alasdair G Kergon Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] device-mapper: Allow referencing by device number Currently userspace code using the dm ioctls must refer to a mapped device by either its name or its uuid. But in some circumstances you know neither of those directly. This patch lets you reference devices by their major/minor numbers as an alternative. Signed-Off-By: Alasdair G Kergon Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Add vram option to intelfb - add vram option to reserve more memory than stolen by BIOS if needed - fix intelfbhw_pan_display typo - add __initdata annotations Signed-off-by: Sylvain Meyer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Fix for using >16 pixel wide font in fb console From: Jani Jaakkola : The first one was that fbcon_set_font() used one byte padding for fonts having width 16 <= width < 24, which was wrong since the pieces of code actually using the font did not use any padding. This is fixed in the included patch and also fbcon_set_font() is made a little cleaner. After the patch the following font is not garbled in fb console: http://www.cs.helsinki.fi/u/jjaakkol/psf/bitstream_vera_sans_mono_roman.... The other bug is that fonts having height == 32 crash the kernel. I have no fix for this (at least not yet), but it can be reproduced with font: http://www.cs.helsinki.fi/u/jjaakkol/psf/bitstream_vera_sans_mono_roman..... Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Support for bigger than 16x32 fonts in softcursor Fix crash if font font is bigger than 16x32 by dynamically allocating buffer based on font dimensions instead of statically allocating at 64 bytes. Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Support for bigger than 16x32 fonts in rivafb cursor - Add support for fonts bigger thatn 16x32 by dynamically allocating buffer based on font dimensions instead of statically allocating at 64 bytes. - use softcursor if cursor size exceeds 32x32. - fix rivafb_cursor if cursor width is not divisible by 2 Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbcon: Disable fbcon cursor if vt softcursor is enabled Problem reported by Gerd Knorr: (1) boot with vesafb (thats what I'm using, maybe it shows on other framebuffers and/or vgacon as well). (2) login into one terminal, then type "echo -ne '\033[?17;15;239c'". You should have a nice, yellow and *not* blinking cursor block. That is what I have in my .profile because I can't stand the blinking cursors. (3) Switch to another terminal. The cursor goes into blinking underscore mode now (i.e. the default cursor). (4) Switch back to the first terminal. Now you have a yellow block with the last two pixel lines (i.e. the underscore) blinking. This bug is caused by both fbcon_cursor and vt softcursor being active at the same time. Fix: - Disable fbcon_cursor if vt softcursor is active (vc->vc_cursor_type & 0x10) != 0. - Recheck/reload fbcon cursor for each vt switch Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Allow mode change even if EDID block is not found - use symbol_get() to check for the presence of i2c - allow mode changing even if EDID block is not found (no I2C support) - fix crashes on chipsets that do not have i2c support - shorten string in info->fix.id - trivial chip name changes Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Fix cursor in doublescan mode in atyfb fix hw cursor in doublescan modes Signed-off-by: Alexander Kern Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Fix typo in atyfb Fix typo, and decrease amount of output Signed-off-by: Alexander Kern Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fbdev: Change the find_mode behavior let find_mode search for nearest refresh rate Signed-off-by: Alexander Kern Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] i2o: changed code with BUG() to BUG_ON() - changed code with BUG() to use BUG_ON() which could be optimized by some platforms (original from Milton Miller) Signed-off-by: Markus Lidel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] i2o: remove unused code and make needlessly global code static - remove unused code - make needlessly global code static Signed-off-by: Adrian Bunk Signed-off-by: Markus Lidel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] i2o: changed old queueing code with wait_event API - removed old queueing code and replaced it with new wait_event API Signed-off-by: Markus Lidel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] i2o: converted SPIN_LOCK_UNLOCKED into spin_lock_init() - changed initialization of spin locks from SPIN_LOCK_UNLOCKED into spin_lock_init() Signed-off-by: Markus Lidel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Remove duplicate safe_for_read(READ_BUFFER) entry in scsi_ioctl.c Signed-off-by: Miles Bader Cc: James Bottomley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Fix badness in scsi_lib.c From: Mike Christie > Oct 26 23:32:55 mina kernel: Unable to handle kernel paging request at > virtual address 6b6b6c7b > Oct 26 23:32:55 mina kernel: printing eip: > Oct 26 23:32:55 mina kernel: f882b8ce > Oct 26 23:32:55 mina kernel: *pde = 00000000 > Oct 26 23:32:55 mina kernel: Oops: 0000 [#1] > Oct 26 23:32:55 mina kernel: PREEMPT > Oct 26 23:32:55 mina kernel: Modules linked in: sd_mod usb_storage > ide_cd cdrom sg scsi_mod rd > Oct 26 23:32:55 mina kernel: CPU: 0 > Oct 26 23:32:55 mina kernel: EIP: 0060:[] Not tainted VLI > Oct 26 23:32:55 mina kernel: EFLAGS: 00010296 (2.6.10-rc1-mm1y) > Oct 26 23:32:55 mina kernel: EIP is at > scsi_block_when_processing_errors+0xe/0xe0 [scsi_mod] > Oct 26 23:32:55 mina kernel: eax: 00000000 ebx: 6b6b6b6b ecx: > f88ef640 edx: ec5b6578 > Oct 26 23:32:55 mina kernel: esi: e9baa4b8 edi: c17e9268 ebp: > e9677f0c esp: e9677eb4 > Oct 26 23:32:55 mina kernel: ds: 007b es: 007b ss: 0068 > Oct 26 23:32:55 mina kernel: Process fdisk (pid: 2891, > threadinfo=e9676000 task=ea0c61f0) > Oct 26 23:32:55 mina kernel: Stack: 00000000 00000001 00000000 00000000 > 00000000 00000000 00000000 00000000 > Oct 26 23:32:55 mina kernel: 00000000 00000000 00000003 e9677ef0 > c17e9268 0000006b c17e9268 e9677f0c > Oct 26 23:32:55 mina kernel: c0159761 c17e93e4 00000000 e9b98780 > e9baa4b8 c17e9268 e9677f24 f88ef6a8 > Oct 26 23:32:55 mina kernel: Call Trace: > Oct 26 23:32:55 mina kernel: [] show_stack+0x7f/0xa0 > Oct 26 23:32:55 mina kernel: [] show_registers+0x156/0x1c0 > Oct 26 23:32:55 mina kernel: [] die+0x156/0x2e0 > Oct 26 23:32:55 mina kernel: [] do_page_fault+0x36d/0x69c > Oct 26 23:32:55 mina kernel: [] error_code+0x2d/0x38 > Oct 26 23:32:55 mina kernel: [] sd_release+0x68/0xa0 [sd_mod] > Oct 26 23:32:55 mina kernel: [] blkdev_put+0x183/0x1b0 > Oct 26 23:32:55 mina kernel: [] __fput+0x14d/0x160 > Oct 26 23:32:55 mina kernel: [] filp_close+0x57/0x90 > Oct 26 23:32:55 mina kernel: [] sys_close+0x113/0x240 > Oct 26 23:32:55 mina kernel: [] sysenter_past_esp+0x52/0x71 > Oct 26 23:32:55 mina kernel: Code: 0c 8b 43 04 8b 00 89 5c 24 04 c7 04 > 24 e4 d1 83 f8 89 44 24 08 e8 d3 21 8f c7 8d 76 00 55 89 e5 57 56 53 83 > ec 4c 8b 75 08 8b 1e <8b> 83 10 01 00 00 a8 08 74 7c fc 31 c0 8d 7d b4 > b9 05 00 00 00 The problem with using shost_for_each_device wrt to the above oops is that scsi_forget_host sets the state to SDEV_CANCEL, so that when scsi_host_cancel iterates over the devices using shost_for_each_device it cannot get a handle to the sdev (scsi_device_get fails becuase the state is set to SDEV_CANCEL). And, __scsi_iterate_devices does not clear the next pointer if this happens, so I think this is needed to fix just the refcount bug in shost_for_each_device. Signed-off-by: James Bottomley [ARM PATCH] 2248/1: S3C2410 - missing serial config in arch/arm/mach-s3c2410/Kconfig Patch from Ben Dooks Add patch that should have been in 2247/1, defining CONFIG_S3C2410_LOWLEVEL_UART_PORT Signed-off-by: Ben Dooks Signed-off-by: Russell King RS library spelling fixes. Originally from Joe Perches Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse MTD: Intel flash chip driver locking fixes. Cleanups and fixes in preparation for XIP support -- fix unbalanced use of get_chip() and put_chip(), and clean up the code a little in preparation for what's to come. Signed-off-by: Nicolas Pitre Signed-off-by: David Woodhouse MTD: DiskOnChip drivers should no longer require old docecc code. Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse MTD map/device driver cleanups -- remove bogus __iomem casts Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse MTD: Fix suspend/resume on Intel flash chip driver We weren't correctly resetting our idea of the chip's state on suspend in all circumstances. Fix that. Signed-off-by: David Woodhouse MTD NAND drivers: cleanup MODULE_PARAM and bogus __iomem casts. Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse MTD: DiskOnChip driver fixes: MODULE_PARAM and __iomem, and fix RS init Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse MTD: Fix oops after erase in NFTL/INFTL (DiskOnChip translation layers) Fix kernel NULL pointer dereference at mtd_erase_callback+6 when trying to insmod INFTL or NFTL modules. Signed-off-by: Kalev Lember Signed-off-by: David Woodhouse MTD: Fix Pb1550 board NAND driver to not read write-only registers Signed-off-by: Pete Popov Signed-off-by: David Woodhouse MTD: Compile fix for the typo fix in ixp2000 map driver. Signed-off-by: Deepak Saxena Signed-off-by: David Woodhouse MTD: Avoid false positives in CFI probe due to floating data bus Signed-off-by: Russell King Signed-off-by: David Woodhouse MTD: Fix detection of hardware partitions in Intel flash chips Determine it from the CFI query data instead of hard-coding it. Signed-off-by: Nicolas Pitre Signed-off-by: David Woodhouse [PATCH] sg: Fix oops of sg_cmd_done and sg_release race The following patch fixes a race condition in sg of sg_cmd_done racing with sg_release. I've seen this bug hit several times on test machines and the following patch fixes it. The race is that if srp->done is set and the waiting thread gets a spurious wakeup immediately afterwards, then the waiting thread can end up executing and completing, then getting closed, freeing sfp before the wake_up_interruptible is called, which then will result in an oops. The oops is fixed by locking around the setting srp->done to 1 and the wake_up, and also locking around the checking of srp->done, which guarantees that the wake_up_interruptible will always occur before the sleeping thread gets a chance to run. Signed-off-by: Brian King Signed-off-by: James Bottomley [PATCH] sym53c500_cs driver update The attached minor patch to linux/drivers/scsi/pcmcia/sym53c500_cs.c allows interrupt sharing, which is evidently a "must have" feature for at least G4 PowerBooks (ppc architecture). The other user of the New Media Bus Toaster reports that his powerbook consistently assigns the yenta CardBus controller IRQ to whatever card he inserts. Signed-off-by: Bob Tracy Signed-off-by: James Bottomley [PATCH] SCSI_QLOGIC_1280_1040 dependencies On Sat, Oct 23, 2004 at 11:55:20AM +1000, Eyal Lebedinsky wrote: > Should the second line be indented? > > Qlogic QLA 1240/1x80/1x160 SCSI support (SCSI_QLOGIC_1280) [M/n/?] m > Qlogic QLA 1020/1040 SCSI support (SCSI_QLOGIC_1280_1040) [N/y/?] (NEW) y Correct, SCSI_QLOGIC_1280_1040 should depend on SCSI_QLOGIC_1280. The patch below does this. Additionally, it prevents the case that both drivers are built statically and both support the 1020/1040. Signed-off-by: Adrian Bunk Signed-off-by: James Bottomley [PATCH] 2.6 aacraid: Interrupt function cleanup This patch updates the adapter health check function and cleans up some unused interrupt related functions and macros. Signed-off-by: Mark Haverkamp Signed-off-by: James Bottomley JFFS2: locking fixes We weren't correctly holding a lock when adding raw nodes to the per-inode list. Signed-off-by: Artem Bityuckiy Signed-off-by: David Woodhouse [PATCH] "mt-st tell" fails in 2.6.10-rc1 On Tue, 2 Nov 2004, Marc Thomas wrote: > I've noticed that in 2.6.10-rc1 the "mt tell" command (from mt-st) fails to > report SCSI tape position, returning "No such device". Other tape (read) > operations appear to be ok. It worked fine in 2.6.9-rc4. The patch at the end of this message should fix this (passes basic tests). Someone added the ioctl to reset the SCSI device but the error handling is not quite correct. The code in 2.6.10-rc1 allows st_ioctl to continue unless reset succeeded. The fix allows continuation only if scsi_nonblockable_ioctl() does not handle the ioctl (returns -ENODEV). retval is cleared because the following code assumes it is zero unless set to something else. Signed-off-by: Kai Makisara Signed-off-by: James Bottomley JFFS2: Remove redundant 'ino' arg from jffs2_get_inode_nodes() We had it already in the inocache structure we were being passed Signed-off-by: Artem Bityuckiy Signed-off-by: David Woodhouse JFFS2: make sync() actually work by providing ->sync_fs method. Signed-off-by: David Woodhouse [PATCH] 2.6 aacraid: rx check health function update This patch updates the rx check health function. My previous only updated the rkt check health code. This and the previous patch should fix the kmalloc return checking bug: http://bugme.osdl.org/show_bug.cgi?id=3699 Signed-off-by: Mark Haverkamp Signed-off-by: James Bottomley JFFS2: jffs2_fs_i.h needs ... and should include it directly rather than hoping it's been done. Signed-off-by: David Woodhouse [PATCH] megaraid 2.20.4.1 Driver From: "Bagalkote, Sreenivas" We are releasing megaraid 2.20.4.1 driver. This version fixes two issues - - Handle IOCTL command timeouts properly - Replace incorrectly introduced pci_dma_sync_{sg,single}_for_cpu with correct pci_dma_sync_{sg,single}_for_device. Signed-off-by: Andrew Morton Signed-off-by: James Bottomley Remove duplicate safe_for_read(READ_BUFFER) From: Miles Bader Signed-off-by: Miles Bader Signed-off-by: Andrew Morton Signed-off-by: James Bottomley [PATCH] Documentation/kernel-parameters.txt: scsi param updates This patch updates Documentation/kernel-parameters.txt with the proper SCSI LUNs params and adds a description for 'max_luns'. Signed-off-by: Cal Peake Signed-off-by: James Bottomley [PATCH] tmascsim: (resend updated) track_queue_full Use scsi_track_queue_full(), don't lie about driver's queueing abilities. Signed-off-by: Guennadi Liakhovetski Signed-off-by: James Bottomley [PATCH] sd.c: adjust READ_CAPACITY for broken devices A number of USB mass storage devices return the wrong value in response to READ CAPACITY: They give the total number of blocks rather than the highest block number. Currently the usb-storage driver tries to adjust for this by altering the contents of the reply buffer. That's not a very good idea because it prevents users from seeing what the device actually has to say; it violates the principle that a transport should do nothing more than transfer data. This patch moves the responsibility for adjusting the capacity to the sd driver instead, by adding a new device flag. After this has been merged the corresponding change (to set the flag) will be made to the usb-storage driver. Signed-off-by: Alan Stern Signed-off-by: James Bottomley [PATCH] refactor tmscsim inititalization code Streamline EEPROM handling code The patch below is a fixed (and slightly reworked) patch from Christoph. Thanks to Chiaki Ishikawa for fixes and tests. Original author: Christoph Hellwig Signed-off-by: Guennadi Liakhovetski Signed-off-by: James Bottomley [PATCH] SCSI: aha1542.c: make some code static The patch below makes some needlessly global code static. Signed-off-by: Adrian Bunk Signed-off-by: James Bottomley [PATCH] SCSI atp870u.c: make a needlessly global function The patch below makes the function is870 which has no external users static. Signed-off-by: Adrian Bunk Signed-off-by: James Bottomley [PATCH] SCSI dc395x.c: make a function static The patch below makes the needlessly global function adapter_init_chip static. Signed-off-by: Adrian Bunk Signed-off-by: James Bottomley [PATCH] SCSI: fdomain.c: make a struct static The patch below makes a needlessly global struct static. Signed-off-by: Adrian Bunk Signed-off-by: James Bottomley JFFS2: fix printk argument type warning fs/jffs2/gc.c:832: warning: signed size_t format, different type arg (arg 3) Signed-off-by: Randy Dunlap Signed-off-by: David Woodhouse [PATCH] SCSI qla1280: some firmware files cleanups - make all this needlessly global code static - remove the unused firmware_version variables Signed-off-by: Adrian Bunk Signed-off-by: James Bottomley [PATCH] SCSI dc395x.c: store pci device pointer Store the PCI device pointer into the adapter control block. This is used in the pci_map_*/pci_unmap_* calls and previously it would have been set to NULL. This appears to be no problem for x86 but is a problem for sparc64. Signed-off-by: Jamie Lenehan Signed-off-by: James Bottomley [PATCH] SCSI dc395x.c: Fix type for irq and io ports Store the port and irq in unsigned long and unsigned int's instead of as u16's and u8's. The later was fine on x86 but not on sparc64. Signed-off-by: Jamie Lenehan Signed-off-by: James Bottomley [PATCH] SCSI dc395x.c: Call pci_disable during cleanup. Call pci_disable_device if initialisation fails and during cleanup. Signed-off-by: Jamie Lenehan Signed-off-by: James Bottomley [PATCH] net: scheduling policing fix Missing goto. Acked by Thomas Graf MTD: some cleanups The patch below makes the following cleanups for code under drivers/mtd/ : - make some needlessly global code static Signed-off-by: Adrian Bunk Signed-off-by: David Woodhouse MTD: Fix memory leak in FTL translation layer. Signed-off-by: David Woodhouse [IA64] fix phys. address conversion in ia64_pal_tr_read The palinfo module may call ia64_pal_tr_read using an IPI. If the processor receiving the IPI is idle, the call will fail since the idle process stack is in region 5 and ia64_pal_tr_read uses __pa rather than ia64_tpa. This small patch fixes the problem by making it use ia64_tpa instead. Signed-off-by: Jesse Barnes Signed-off-by: Tony Luck [IA64] remove dead wood from asm-ia64/hardirq.h syscall_count, ksoftirqd_task, local_syscall_count, local_ksoftirqd_task and local_nmi_count are totally unused (and gone from other architectures aswell) nmi_count() is a stub only used once in arch/ia64/kernel/irq.c, so we can use 0 directly there. I suspect it was added to keep irq.c in sync with i386, but even the CONFIG_GENERIC_HARDIRQS code (which IA64 needs to be converted to) leaves /proc/interrupts output to the architecture because it's so different over the architectures. Signed-off-by: Christoph Hellwig Deleted the code to print "NMI:" and a row of zeroes from irq.c -Tony Signed-off-by: Tony Luck SCSI: Fix Bug 3753 (multiple definition of ST_partstat) Fix a missed piece of junk from a typedef conversion in st.h Signed-off-by: James Bottomley [PATCH] Lock initializer unifying Batch 2 (SCSI) To make spinlock/rwlock initialization consistent all over the kernel, this patch converts explicit lock-initializers into spin_lock_init() and rwlock_init() calls. Currently, spinlocks and rwlocks are initialized in two different ways: lock = SPIN_LOCK_UNLOCKED spin_lock_init(&lock) rwlock = RW_LOCK_UNLOCKED rwlock_init(&rwlock) this patch converts all explicit lock initializations to spin_lock_init() or rwlock_init(). (Besides consistency this also helps automatic lock validators and debugging code.) The conversion was done with a script, it was verified manually and it was reviewed, compiled and tested as far as possible on x86, ARM, PPC. There is no runtime overhead or actual code change resulting out of this patch, because spin_lock_init() and rwlock_init() are macros and are thus equivalent to the explicit initialization method. That's the second batch of the unifying patches. Signed-off-by: Thomas Gleixner Acked-by: Ingo Molnar Signed-off-by: James Bottomley [ARM PATCH] 2251/1: S3C2410 - system timer rename [cosmetic] Patch from Ben Dooks Correct the name used when registering the system tick interrupt from S32410 to S3C2410. Signed-off-by: Dimitry Andric Signed-off-by: Ben Dooks Signed-off-by: Russell King [libata ahci] minor fixes Add support for ioctl handling. Remove incorrect comment. [PATCH] io context leak on queue drain If the drain flag is set on the queue, get_request() needs to drop the acquired iocontext to avoid leaks. Signed-off-by: Jens Axboe Signed-off-by: Linus Torvalds [PATCH] Fork fix fix The attached patch fixes the fork fix to avoid the divide-by-zero error I'd previously fixed, but without using any sort of conditional. Signed-off-by: David Howells Signed-off-by: Linus Torvalds [PATCH] binfmt_elf: handle p_filesz == 0 on PT_INTERP section Jakub Jelinek points out that current fix has an underflow problem if elf_ppnt->p_filesz == 0. Fix that up, and also stop overwriting interpreter buffer, simply check that it's NULL-terminated. From: Jakub Jelinek Signed-off-by: Chris Wright Signed-off-by: Linus Torvalds [PATCH] setup_arg_pages can insert overlapping vma Florian Heinz built an a.out binary that could map bss from 0x0 to 0xc0000000, and setup_arg_pages() would be unhappt in insert_vma_struct because the arg pages overlapped. This just checks before inserting, and bails out if it would overlap. Signed-off-by: Chris Wright Signed-off-by: Linus Torvalds [PATCH] a.out: error check on set_brk It's possible for do_brk() to fail during set_brk() when exec'ing and a.out. This was noted with Florian's a.out binary and overcommit set to 0. Capture this error and terminate properly. Signed-off-by: Chris Wright Signed-off-by: Linus Torvalds [SPARC64]: Couple of do_sparc64_fault fixes. 1) Use in_atomic() for atomicity check. 2) Add mmap_sem deadlock on bad kernel access prevention just as on i386 3) Use VM_FAULT_* macros instead of magic constants. Signed-off-by: David S. Miller [SPARC64]: Two io_remap_page_range() fixes. 1) BUG on presence of existing mappings just as remap_pfn_range does 2) Perform TLB flush while holding page_table_lock Signed-off-by: David S. Miller [SPARC64]: Do not set VM_LOCKED on I/O mapped areas. Use VM_RESERVED instead. Signed-off-by: David S. Miller [SPARC64]: Update defconfig. Signed-off-by: David S. Miller [Bluetooth] Update copyright information The initial firmware loading routine for the Broadcom based chips was written by Maxim Krasnyansky and so he also owns a copyright for this driver. Signed-off-by: Marcel Holtmann [Bluetooth] Correct locking for zero SCID responses The l2cap_get_chan_by_ident() function must lock the socket with the bh_lock_sock() function on success. Signed-off-by: Marcel Holtmann [Bluetooth] Add support for L2CAP secure mode The secure mode requests an authentication and enables the encryption for that connection. After the successful execution of these procedures it triggers the generation of a new link key. Signed-off-by: Marcel Holtmann [Bluetooth] Check for L2CAP reliability The L2CAP sockets can now set LM_RELIABLE flag and get notification when a reliability problem with the ACL connection is detected. The Bluetooth qualification tests require this functionality. Signed-off-by: Maxim Krasnyansky Signed-off-by: Marcel Holtmann [PATCH] x86-64: fix boot crash on VIA systems quirk_via_irqpic cannot be __devinit because it runs at pci_enable_device() time now. This fixes a boot time crash on a VIA x86-64 machine. Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds [PATCH] Make dss1_divert ISDN module work on SMP again When I switched my installation from kernel 2.4 to 2.6 I recognized that the ISDN module dss1_divert was marked incompilable (config option CONFIG_CLEAN_COMPILE must be turned off). The compile problem was the obsolete using of kernel 2.4 critical sections. I replaced the cli() stuff with spinlocks as explained in the Documentation/spinlocks.txt file. After that the module compiles and runs as expected. Signed-off-by: Steffen A. Mork Signed-off-by: Linus Torvalds [libata docs] add chapter on libata driver API [ide] update CRISv10 IDE driver Signed-off-by: Mikael Starvik Signed-off-by: Bartlomiej Zolnierkiewicz [ide] no need to alloc sg_table in CRISv10 IDE driver Signed-off-by: Bartlomiej Zolnierkiewicz [ide] update documentation for ide params From: Magnus Damm This patch removes ide parameters marked as obsolete in the source and adds documentation for "ide=". I think I got it right, the important part for me is "ide=nodma". Signed-off-by: Bartlomiej Zolnierkiewicz [ide] "ide=nodma" printout fix From: Magnus Damm This simple patch changes the output from this: .. ide_setup: ide=nodmaIDE: Prevented DMA .. to this: .. ide_setup: ide=nodma : Prevented DMA .. Signed-off-by: Bartlomiej Zolnierkiewicz [ide] remove RICOH CD-R/RW MP7083A from DMA blacklist From: Srihari Vijayaraghavan I have tested my RICOH CD-R/RW with this patch (on CD reading/writing), and it works just fine with DMA enabled. Signed-off-by: Bartlomiej Zolnierkiewicz [ide] small IDE cleanups The patch below does the following small cleanups in the IDE code: - make some needlessly global code static - remove two unused functions from pdc202xx_new.c Signed-off-by: Adrian Bunk Signed-off-by: Bartlomiej Zolnierkiewicz [ide] fix /proc/ide/hd?/settings to not spam logs On Thu, 11 Nov 2004 17:10:21 +0100, Jens Axboe wrote: > On Thu, Nov 11 2004, Alan Cox wrote: > > > + printk(KERN_WARNING "Warning: /proc/ide/hd?/settings interface is " > > > + "obsolete, and will be removed soon!\n"); > > > + > > > > The above should be rate limited or on the write case moved to after > > the capable() check. A program polling these settings now makes a nasty > > noise and wipes the logs. A user can also do it intentionally. > > Or just print it once... Signed-off-by: Bartlomiej Zolnierkiewicz [PATCH] fix dss1_divert fixes As noted by Bartlomiej, this makes the spinlock be a proper lock rather than just a local lock to each user. Signed-off-by: Steffen A. Mork Signed-off-by: Linus Torvalds [IPV4/IPV6]: Remove frag_list check from output path. I've removed the frag_list check before fragmenting and we'll rely on dev_queue_xmit to fix things up if necessary. Any functions in between should do the right thing since they need to handle non-linear skb's anyway. Well that's the theory :) Signed-off-by: Herbert Xu Signed-off-by: David S. Miller [IPV6]: Set sk_prot early enough in inet6_create() It needs to be before the sk_set_owner() call else we OOPS. Signed-off-by: David S. Miller Fix floppy driver lock-up when you have an irq storm. It happens with misconfigured interrupts (and no floppy controller). We should probably also make floppy use the irq probing mechanism, which handles that situation more gracefully. [PATCH] rivafb: fix broken burst length calculation As noted by Linus, the burst-length "log2()" calculations were broken by the cleanup. This reverts to the old code. [PATCH] fbdev: Fix rivafb breakage (typo introduced by NV IO access cleanups) Fix typo introduced during changes to use NV_ macros, which caused screen corruption when using rivafb Signed-off-by: Andrew Walrond Signed-off-by: Linus Torvalds Fix reading /proc//mem when parent dies. We should not touch "self_exec_id" here. The parent changed, not we. [IA64] perfmon: enable interrupts around semaphore call fix a problem in pfm_handle_work() where a task may go to sleep in down_interruptible() with interrupts disabled. This may happen only for a monitored thread for which the PFM_FL_NOTIFY_BLOCK flag is set. The fix re-enables interrupts around the semaphore call. It is safe to do so because pfm_handle_work() is called from do_notify_user() which is called only when the task is about to leave the kernel, i.e., no risk of stack overflow because of interrupt nesting. fix a debug printk warning in pfm_write_pmcs() signed-off-by: Stephane Eranian ack-by: David Mosberger Signed-off-by: Tony Luck [DECNET]: Remove DECNET_SIOCGIFCONF. It's only purpose seems to be to break ifconfig. It often gets enabled by default by distributors who like to set everything on, and it just confuses people. The functionality is available via netlink anyway - which is what anyone using DECnet will be using for configuration. Signed-off-by: Patrick Caulfield Signed-off-by: David S. Miller [SCTP]: Fix printk arg type fix printk argument type warning: net/sctp/socket.c:2672: warning: format argument is not a pointer (arg 5) Signed-off-by: Randy Dunlap Signed-off-by: David S. Miller [IPV4]: Missing pskb_may_pull in icmp_filter. Signed-off-by: David S. Miller [VLAN]: change_mtu should return 0 on success. Signed-off-by: John W. Linville Signed-off-by: David S. Miller [ARM PATCH] 2249/1: S3C2410 - update help for arch/arm/Kconfig.debug Patch from Ben Dooks Update help-text to show the new option for controlling where the low-level debug messgaes go. Signed-off-by: Ben Dooks Signed-off-by: Russell King [ARM PATCH] 2256/1: S3C2410 - dma load fixes Patch from Ben Dooks Fix pair of bugs in dma loading code 1) no need to error if loading onto stopped channel 2) irq code prints error if buffer done callback fully reloads the channel. Signed-off-by: Ben Dooks Signed-off-by: Russell King [RANDOM]: Remove TCP MD4 code if not CONFIG_INET Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller [AF_UNIX]: Fix SELinux crashes with SOCK_SEQPACKET 1) Don't call security_unix_may_send() hook during sendmsg() for SOCK_SEQPACKET, and ensure that sendmsg() can only be called on a connected socket so as not to bypass the security_unix_stream_connect() hook. 2) Return -EINVAL if sendto() is called on SOCK_SEQPACKET with an address supplied. Signed-off-by: James Morris Signed-off-by: David S. Miller [ARM PATCH] 2263/1: S3C2410 - gpio pin config fixes Patch from Ben Dooks Fix S3C2410_GPE3, S3C2410_GPE4 and S3C2410_GPE7 pin functions, and add S3C2440 AC97 pin functions Signed-off-by: Ben Dooks Signed-off-by: Russell King [ARM PATCH] 2252/1: Fix IXP4XX timer interrupt implementation Patch from Deepak Saxena The current timer interrupt implementation can cause time to skip forward by ~65s and causes a 1 minute pause during bootup. The fix for this was found during 2.4 but got lost in the 2.6 transition. Details @: http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2003-September/... Signed-off-by: Deepak Saxena Signed-off-by: Russell King [ARM PATCH] 2253/1: Fix IXP4xx PCI config cycle routines Patch from Deepak Saxena The IXP4xx PCI config cycle routines currently virtualize accesses to device 0:0 to map to the host bridge itself. This is technically incorrect b/c on certain boards we have an actual device wired to 0:0 and the existing code will not see these. Signed-off-by: Deepak Saxena Signed-off-by: Russell King SCSI: fix USB forced remove oops Because of the changes to add the target in to the driver model, the cancellation method no-longer works correctly. Fix it by iterating using shost_for_each_device instead. Signed-off-by: James Bottomley [ARM PATCH] 2254/1: Fix ixp4xx-regs.h PCI config address typo Patch from Deepak Saxena There seems to be a typo that has creeped into include/asm/arch-ixp4xx/ixp4xx-regs.h. IXP4XX_EXP_CFG_BASE_VIRT and IXP4XX_PCI_CFG_BASE_VIRT were both defined to the same value. The patch changes PCI_CFG_BASE_VIRT to the value in the memory map comment. Signed-off-by: Deepak Saxena Signed-off-by: Russell King [ARM PATCH] 2258/1: Add missing IXP2000 Makefile.boot file Patch from Deepak Saxena Signed-off-by: Deepak Saxena Signed-off-by: Russell King [ARM PATCH] 2262/1: Various IXP2000 typo fixes and comment cleanups Patch from Deepak Saxena Patch from Lennert Buytenhek Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King [ARM PATCH] 2268/1: Update Documentation/arm/Booting Patch from Deepak Saxena One of the problems I have seen with several bootloaders is that they forget to disable network devices before jumping into the kernel and we end up with kernel text being overwritten by random data before the in-kernel driver resets the devices. This patch adds a blurb to the ARM Bootloader documentation reminding developers to quiesce all DMA capable devices before executing the kernel. Signed-off-by: Deepak Saxena Signed-off-by: Russell King [SPARC]: Fix serial console handling. Due to recent logic changes in uart_add_one_port(), the serial console was being registered twice. To avoid this, so serial console init before we actually register the uarts ports with the core serial layer. Noticed by William Lee Irwin III Signed-off-by: David S. Miller [IA64] perfmon: fix double end-of-comment in previous checkin Following trivial patch fixes what looks like a merge bug. Signed-off-by: Grant Grundler Signed-off-by: Tony Luck [PCMCIA] Don't place Cardbus bridges into D3 state on suspend. Some laptops, notably IBM T22, refuse to suspend when the Cardbus bridge is placed into D3. Therefore, comment out the call for the time being. [PATCH] fix kobject varargs bug From: Gerd Knorr It uses the varargs list twice in a illegal way. That doesn't harm on i386 by pure luck, but blows things up on amd64 machines. Using var args list twice without calling va_start twice is illegal. Signed-off-by: Gerd Knorr Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman [PATCH] fix oops in sysfs_remove_dir() The following patch should avoid the sysfs_remove_dir() oops you are seeing while device removal. It anyway fixes the obvious error and is needed. But it will not make any change to the first error you are seeing while connecting the device. o Following patch avoids the sysfs_remove_dir() oops when it is passed a kobject with NULL dentry. Signed-off-by: Maneesh Soni Signed-off-by: Greg Kroah-Hartman [PATCH] I2C: Do not register useless smsc47m1 While verifying my stack of patches against what you sent to Linus last week, I noticed this one. Looks like I simply forgot to send it to you, as I cannot find any trace of it in the lm_sensors mailing-list archives. The patch prevents an smsc47m1 device from being registered when no monitoring function is actually active within the chip. See this ticket for background: http://secure.netroedge.com/~lm78/readticket.cgi?ticket=1801 This is certainly better to explicitely fail in this case than leave the user with an empty sysfs directory (except for alarms), which tends to make him/her think of a driver bug, which it isn't (what it really is is a BIOS brokenness). Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman [PATCH] I2C: Fixes to the i2c-amd756-s4882 driver While working on the 2.4 version of the i2c-amd756-s4882 driver, I noticed a few quirks on the 2.6 version I sent to you. The following patch attempts to fix them. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman [PATCH] I2C: Cleanups to the recent smbus functions removal This patch cleans up the recent removal of smbus functions proposed by Arjan and then fixed by Gabriel. Changes are as follow: 1* Discard i2c_smbus_block_process_call, as it isn't used anywhere either. I guess that Arjan missed it because it wasn't exported. 2* Document the functions removal, so that people have at least an idea that the functions can be restored later if needed. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman [PATCH] I2C: minor comment fix It seems so. BTW I hate wrong comments and happened to add one in my patch. To fix my blunder, can you apply the appended one line removal on top of Jean's patch. Signed-off-by: Gabriel Paubert Signed-off-by: Greg Kroah-Hartman [PATCH] I2C: i2c-nforce2.c add support for nForce3 Pro 150 MCP This is the all new and improved version of the patch: - following the advise from Jean Delvare I removed the redundant definition of the PCI IDs from the driver and just add them to the pci_ids.h file. - the patch is now created against linux 2.6.10-RC2. Signed-off-by: Thomas Leibold Signed-off-by: Greg Kroah-Hartman [PATCH] USB: Add some help text for touchkitusb add some help text for touchkitusb. the howto there should also work for mtouchusb and other touchscreens using the input layer. Signed-off-by: Daniel Ritz Signed-off-by: Greg Kroah-Hartman [PATCH] USB: pegasus endian fixes big-endian related fixes; Signed-off-by: Petko Manolov Signed-off-by: Greg Kroah-Hartman [PATCH] USB: add PID to ftdi_sio.c The attached patch fixes a bug introduced by myself (a brown paper bag one) when I posted the patch that introduced this same PID to the id_table_FT232BM array only and not the id_table_combined array. Signed-off-by: Paulo Marques Signed-off-by: Greg Kroah-Hartman [PATCH] USB: visor: Always do generic_startup generic_startup in visor.c was not called for some hardware, resulting in attempts to access memory that had never been allocated, which in turn caused the problem several people reported with recent (2.6.10ish) kernels. Signed-off-by: Roger Luethi Signed-off-by: Greg Kroah-Hartman [PATCH] Re: The "ipaq" module: Updated list of vendor/product IDs This patch for ipaq.h and ipaq.c in drivers/usb/serial/ has the following features: o Provides an updated table with all vendor/product IDs (except one [1]) present in Microsoft ActiveSync 3.7.1, the latest version available for download. o Removes #defines with vendor and product IDs from ipaq.h [2] The table has been autogenerated from the file wceusbsh.inf (extracted from MSASYNC.EXE with cabextract). The scripts used to generate the table become available for download as part of the SynCE project (http://synce.sourceforge.net). [1] Vendor 0x45e (Microsoft) with device 0xCE is not included, because they are used when the device is turned off but still connected to the PC via USB. When the device is turned on, it changes vendor/product IDs to the more specific ones. [2] Not needed with an autogenerated device table Signed-off-by: David Eriksson Signed-off-by: Ganesh Varadarajan Signed-off-by: Greg Kroah-Hartman [PATCH] USB Storage: Unusual_dev entry for tekom/yakumo Here's another unusual_devs entry for the Tekom/Yakumo devices. While I was there I realized two of the tekom entries were out of order, so I fixed that as well. From: Paul Ortyl Signed-off-by: Phil Dibowitz Signed-off-by: Greg Kroah-Hartman [PATCH] USB: new defice for usb serial pl2303 Add device ID to pl2303 usb serial driver for 1) I/O DATA USB-RSAQ3 2) ELECOM UC-SGT Signed-off-by: CHIKAMA Masaki Signed-off-by: Greg Kroah-Hartman [PATCH] USB Storage: fixes to usb-storage scanning thread This was originally patch as401 from Alan Stern. This patch makes a couple of small changes to the new scanning thread. It changes the name to "usb-stor-scan", which was my original intention and makes more sense than the current name. It also removes the PF_NOFREEZE flag; if someone suspends their system immediately after plugging in a USB storage device then the scanning thread should be frozen along with everything else. The delay will be restarted after the system wakes up. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman [PATCH] USB Storage: Force INQUIRY length to be 36 This is patch as399, originally from Alan Stern. A recent change to the scsi_probe_lun() routine now allows host drivers to specify a device's INQUIRY length in a way that cannot be overridden by the value returned from the device. This patch makes usb-storage set the length to 36; now buggy devices won't be able to cause trouble by saying that they have 37 bytes of INQUIRY data available. The only way this value could be changed is if someone creates a SCSI blacklist entry with the BLIST_INQUIRY_58 flag -- and no one would do that for a USB device (I hope)! Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman [PATCH] USB Storage: Remove unnecessary state testing This patch started as as405 from Alan Stern. It has been re-generated against the current tip of the BK tree. For quite a while we've had a bunch of state-transition testing code in the driver, to report if anything bad ever happens (like the SCSI midlayer trying to queue a second command before the first one finishes). None of those tests triggered in a very long time; this aspect of the code appears to be extremely stable. So this patch removes all those tests for illegal values of us->sm_state. It turns out that sm_state was used only for one other purpose: to check whether a command had timed out and caused a SCSI abort. That piece of information can easily be stored in a single new bitflag (which is called calling US_FLIDX_TIMED_OUT) and doing so makes us->sm_state completely unused. Hence the patch removes it from the structure. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman [PATCH] PCI Hotplug: remove unused drivers/pci/hotplug/pciehp_sysfs.c Remove unused the drivers/pci/hotplug/pciehp_sysfs.c Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman [PATCH] PCI Hotplug: clean up rpaphp_pci.c::rpaphp_find_pci_dev this patch improves rpaphp_find_pci_dev. First it uses the for_each_pci_dev macro instead of the while loop, making this hotplug safe (which is a good idea in a hotplug driver, isn't it?). Then it removes retval_dev. retval_dev is set to the found device when something is found, NULL otherwise. If nothing is found dev will be NULL at the end of the loop anyway and the found device if we found something, no need for retval_dev then. And a very small coding style fix. Signed-off-by: Rolf Eike Beer Signed-off-by: Greg Kroah-Hartman [PATCH] PCI Hotplug: cpcihp_generic: fix module_param data type drivers/pci/hotplug/cpcihp_generic.c:214: warning: return from incompatible pointer type Signed-off-by: Randy Dunlap Signed-off-by: Greg Kroah-Hartman [PATCH] PCI: fix Documentation/pci.txt inconsistency Signed-off-by: Rolf Eike Beer Signed-off-by: Greg Kroah-Hartman [PATCH] x86_64: fix async IPIs Fix async IPI to not return before stack data is used. This can fix reboot crashes and apparently helps vmware. Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds [PATCH] 3ware bad queuecommand returns If your ->queuecommand() completes a command through done, it must not return non-zero to the mid layer so that it takes ownership of the command again. Signed-off-by: Jens Axboe corrected typo Signed-off-by: James Bottomley [PATCH] ide-scsi bad queuecommand return ide-scsi completes a command but returns ownership to the mid layer. Signed-off-by: Jens Axboe Signed-off-by: James Bottomley [PATCH] megaraid bad queuecommand return In case of scb allocation failure, mega_build_cmd() can return NULL scb but set busy, which causes a non-zero return from ->queuecommand() while the scsi command has been completed. Just returning busy without calling done() should be enough for appropriate retries. Signed-off-by: Jens Axboe Signed-off-by: James Bottomley [PATCH] ncr53c8xx bad queuecommand return One more case of returning non-zero with a completed command. Signed-off-by: Jens Axboe Signed-off-by: James Bottomley [PATCH] nsp32 bad queuecommand return The driver appropriately returns SCSI_MLQUEUE_HOST_BUSY, but it has already completed the command. Just return 0 instead and let the mid layer handle it through scsi_done(). Signed-off-by: Jens Axboe Signed-off-by: James Bottomley [PATCH] aacraid bad queuecommand return aac_read() and aac_write() may complete a command but return -1, which is propagated through aac_scsi_cmd() to the mid layer through ->queuecommand. If the command has been completed, it must return 0. Signed-off-by: Jens Axboe Signed-off-by: James Bottomley [PATCH] megaraid2 bad queuecommand return If the command is already completed, megaraid_queue_command() must return 0. Signed-off-by: Jens Axboe Signed-off-by: James Bottomley [PATCH] nsp_cs bad queuecommand return Return 0 from queuecommand, if the command has been completed already. Signed-off-by: Jens Axboe Signed-off-by: James Bottomley [PATCH] selinux: cache not freed if load_policy fails; reload BUG's If security_load_policy() fails on the first try, the cache is never cleaned up. When the policy is fixed and a reload is attempted, the old cache will still exist, causing a BUG() in kmem_cache_create(). This patch adds a destroy operation to clean up the cache on failure. Signed-off-by: Jeff Mahoney Signed-off-by: Linus Torvalds [PATCH] sched: fix ->nr_uninterruptible handling bugs PREEMPT_RT on SMP systems triggered weird (very high) load average values rather easily, which turned out to be a mainline kernel ->nr_uninterruptible handling bug in try_to_wake_up(). the following code: if (old_state == TASK_UNINTERRUPTIBLE) { old_rq->nr_uninterruptible--; potentially executes with old_rq potentially being != rq, and hence updating ->nr_uninterruptible without the lock held. Given a sufficiently concurrent preemption workload the count can get out of whack and updates might get lost, permanently skewing the global count. Nothing except the load-average uses nr_uninterruptible() so this condition can go unnoticed quite easily. the fix is to update ->nr_uninterruptible always on the runqueue where the task currently is. (this is also a tiny performance plus for try_to_wake_up() as a stackslot gets freed up.) while fixing this bug i found three other ->nr_uninterruptible related bugs: - the update should be moved from deactivate_task() into schedule(), beacause e.g. setscheduler() does deactivate_task()+activate_task(), which in turn may result in a -1 counter-skew if setscheduler() is done on a task asynchronously, which task is still on the runqueue but has already set ->state to TASK_UNINTERRUPTIBLE. sys_sched_setscheduler() is used rarely, but the bug is real. (The fix is also a small performance enhancement.) The rules for ->nr_uninterruptible updating are the following: it gets increased by schedule() only, when a task is moved off the runqueue and it has a state of TASK_UNINTERRUPTIBLE. It is decreased by try_to_wake_up(), by the first wakeup that materially changes the state from TASK_UNINTERRUPTIBLE back to TASK_RUNNING, and moves the task to the runqueue. - on CPU-hotplug down we might zap a CPU that has a nonzero counter. Due to the fuzzy nature of the global counter a CPU might hold a nonzero ->nr_uninterruptible count even if it has no tasks anymore. The solution is to 'migrate' the counter to another runqueue. - we should not return negative counter values from the nr_uninterruptible() function, since it accesses them without taking the runqueue locks, so the total sum might be slightly above or slightly below the real count. I tested the attached patch on x86 SMP and it solves the load-average problem. (I have tested CPU_HOTPLUG compilation but not functionality.) I think this is a must-have for 2.6.10, because there are apps that go berzerk if load-average is too high (e.g. sendmail). Signed-off-by: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64 iSeries: PURR emulation fix It appears that the PURR emulation patch was the victim of the renaming of a structure element. The current BK kernel will not build for iSeries without this patch. Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] parport_pc CONFIG_PCI=n build fix The following patch fixes these build errors on machines with CONFIG_PCI=n: drivers/parport/parport_pc.c:3199: error: `parport_init_mode' undeclared (first use in this function) drivers/parport/parport_pc.c:3199: error: (Each undeclared identifier is reported only once drivers/parport/parport_pc.c:3199: error: for each function it appears in.) Signed-off-by: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] parport_pc warning fixes Stomp a couple of warnings which several people have tried to fix. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: Fix build error of arch/m32r/mm/fault.c The modification of "Changes for arch/m32r/mm/fault.c@1.3" was prepared for enforce-a-gap-between-heap-and-stack.patch(*) of -mm tree, but it has not been merged into mainline. (*) "heap-stack-gap for 2.6" (Sep. 25, 2004) http://www.uwsg.iu.edu/hypermail/linux/kernel/0409.3/0435.html So, this patch is for withdrawing the previous arch/m32r/mm/fault.c. Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fix for mpol mm corruption on tmpfs With the inline symlink shmem_inode_info structure is overwritten with data until vfs_inode, and that caused the ->policy to be a corrupted pointer during unlink. It wasn't immediatly easy to see what was going on due the random mm corruption that generated a weird oops, it looked more like a race condition on freed memory at first. There's simply no need to set a policy for inodes, since the idx is always zero. All we have to do is to initialize the data structure (the semaphore may need to run during the page allocation for the non-inline symlink) but we don't need to allocate the rb nodes. This way we don't need to call mpol_free during the destroy_inode (not doable at all if the policy rbtree is corrupt by the inline symlink ;). An equivalent version of this patch based on a 2.6.5 tree with additional numa features on top of this (i.e. interleaved by default, and that's prompted me to add a comment in the LNK init path), works fine in a numa simulation on my laptop (untested on the bare hardware). Signed-off-by: Andrea Arcangeli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] mempolicy can select the wrong policy mempolicy.c code will return the wrong policy in some cases. Signed-off-by: Andrea Arcangeli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] kprobes: dont steal interrupts from vm86 This patch fixes the problem reported by Stas Sergeev, that kprobes steals the virtual-8086 exceptions. This fix modifies kprobe_handler() to return 0 when in virtual-8086 mode. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Fix ia64 flush_tlb_page build error Fix ia64 build error: implicit declaration of function `flush_tlb_page' Signed-off-by: Bjorn Helgaas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] cdrom: handle SYSCTL without PROC_FS Fixes oops (reference to cdrom_root_table->child->) when CONFIG_SYSCTL=y and CONFIG_PROC_FS=n. Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] early uart console support This adds an early polled-mode "uart" console driver, based on Andi Kleen's early_printk work. The difference is that this locates the UART device directly by its MMIO or I/O port address, so we don't have to make assumptions about how ttyS devices will be named. After the normal serial driver starts, we try to locate the matching ttyS device and start a console there. Sample usage: console=uart,io,0x3f8 console=uart,mmio,0xff5e0000,115200n8 If the baud rate isn't specified, we peek at the UART to figure it out. Signed-off-by: Bjorn Helgaas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] move HCDP/PCDP to early uart console This changes the HCDP/PCDP support to use the early uart console rather than using early_serial_setup(). As a consequence, ia64 serial device names will now stay constant regardless of firmware console settings. (A serial device selected as an EFI console device on HP ia64 boxes used to automatically become ttyS0.) This also removes the ia64 early-boot kludge of assuming legacy COM ports at 0x3f8 and 0x2f8. For boxes that have legacy ports but no HCDP, "console=ttyS0" will still work, but the console won't start working until after the serial driver initializes and discovers the devices. WARNING: If you have an HP machine and you're using the MP serial console port (the connector labelled "console" on the 3-headed cable), this patch will break your console! HOW TO FIX IT: 1) The console device will change from /dev/ttyS0 to /dev/ttyS1, ttyS2, or ttyS3, so: 1a) Edit /etc/inittab to add a getty entry for /dev/ttyS1 (rx4640, rx5670, rx7620, rx8620, Superdome), /dev/ttyS2 (rx1600), or /dev/ttyS3 (rx2600). 1b) Edit /etc/securetty to add ttyS1, ttyS2, or ttyS3. 1c) Leave the existing ttyS0 entries in /etc/inittab and /etc/securetty so you can still boot old kernels. 2) Edit /etc/elilo.conf to remove any "console=" arguments (see [1]). 3) Run elilo to install the bootloader with new configuration. 4) Reboot and use the EFI boot option maintenance menu to select exactly one device for console output, input, and standard error. Then do a cold reset so the changes take effect. For the MP console, be careful to select the device with "Acpi(HWP0002,700)/Pci(...)/Uart" in the path (see [2]). DETAILS: - Prior to this patch, serial device names depended on the HCDP, which in turn depends on EFI console settings. After this patch, the naming always stays the same, regardless of firmware settings. For example, an rx1600 with a single built-in serial port plus an MP has these ports: Old Old MMIO (EFI console (EFI console address on builtin) on MP port) New ========== ========== ========== ====== builtin 0xff5e0000 ttyS0 ttyS1 ttyS0 MP UPS 0xf8031000 ttyS1 ttyS2 ttyS1 MP Console 0xf8030000 ttyS2 ttyS0 ttyS2 MP 2 0xf8030010 ttyS3 ttyS3 ttyS3 MP 3 0xf8030038 ttyS4 ttyS4 ttyS4 - If you want to have multiple devices in the EFI console path, you can, but Linux won't be able to deduce which console to use, so it will default to using VGA. You can use "console=hcdp" (the UART device from the EFI path) or "console=ttyS" to select the device directly. TROUBLESHOOTING: - No kernel output after "Uncompressing Linux... done": -> You're using an MP port as the console and specified "console=ttyS0". This port is now named something else. Remove the "console=" option. -> Multiple UARTs selected as EFI console devices, and you're looking at the wrong one. Make sure only one UART is selected (use the EFI Boot Manager "Boot option maintenance" menu). -> You're physically connected to the MP port but have a non-MP UART selected as EFI console device. Either move the console cable to the non-MP UART, or change the EFI console path to the MP UART (the MP UART is the one with "Acpi(HWP0002,700)/Pci(...)/Uart" in it.) - Long pause (60+ seconds) between "Uncompressing Linux... done" and start of kernel output: -> No early console, probably because you used "console=ttyS". Remove the "console=" option. - Kernel and init script output works fine, but no "login:" prompt: -> Add getty entry to /etc/inittab for console tty. Use the table in (1a) above or look for the "Adding console on ttyS" message that tells you which device is the console. - "login:" prompt, but can't login as root: -> Add entry to /etc/securetty for console tty. [1] When the EFI console path contains exactly one device (either serial or VGA), 2.6.6 and newer kernels default to that device automatically. So if you remove "console=" arguments, you can use the same elilo configuration to boot any 2.6.6 or newer kernel with or without this patch. If you need to boot kernels older than 2.6.6 (including RHEL3 and SLES9), keep an 'append="console=ttyS0"' line in those elilo.conf stanzas. Non-HP machines will still need "console=" for serial consoles because they don't supply the HCDP table. [2] The HP management card (MP) causes confusion because it is always active as an EFI console, even if it doesn't appear in the EFI console path. If your console path is set to a non-MP UART, and you happen to be attached to the MP UART, everything works in EFI, but the kernel will think the non-MP UART is the console, so you won't see any kernel output. Signed-off-by: Bjorn Helgaas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] gbefb.c build fix The current gbefb.c source cannot be compiled as module because of a small typo where "option" was written instead of "options" in two places. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] vmscan: ignore swap token when in trouble The token-based thrashing control patches introduced a problem: when a task which doesn't hold the token tries to run direct-reclaim, that task is told that pages which belong to the token-holding mm are referenced, even though they are not. This means that it is possible for a huge number of a non-token-holding mm's pages to be scanned to no effect. Eventually, we give up and go and oom-kill something. So the patch arranges for the thrashing control logic to be defeated if the caller has reached the highest level of scanning priority. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] md: Fix problem with unsigned variable going "negative" in linear.c We replace 'size' by 'start'. 'start' means exactly the same as 'curr_offset - size', and the equivalence of the new code can be tested based on this. The difference is that 'start' will never be negative and so can fit in a 'sector_t' while 'size' could be negative. Also make curr_offset sector_t, as it should have been. Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] mlock-vs-VM_IO hang fix With Andrea Arcangeli Fix a hang which occurs when mlock() encounters a mapping of /dev/mem. These have VM_IO set. follow_page() keeps returning zero (not a valid pfn) and handle_mm_fault() keeps on returning VM_FAULT_MINOR (there's a pte there), so get_user_pages() locks up. The patch changes get_user_pages() to just bale out when it hits a VM_IO region. make_pages_present() is taught to ignore the resulting -EFAULT. We still have two bugs: a) If a process has a VM_IO vma, get_user_pages() will bale early, without having considered the vmas at higher virtual addresses. As do_mlock() also walks the vma list this bug is fairly benign, but get_user_pages() is doing the wrong thing there. b) The `len' argument to get_user_pages should be long, not int. We presently have a 16TB limit on 64-bit. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ISDN divert_init.c build fix drivers/isdn/divert/divert_init.c:25: error: conflicting types for 'printk' Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc32: Fix CONFIG_8260 and CONFIG_BLK_DEV_INITRD found that CONFIG_8260 and CONFIG_BLK_DEV_INITRD was broken as arch/ppc/syslib/m8260_setup.c was missing . Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ST_partstat multiple definition drivers/scsi/osst.o(.bss+0x0): multiple definition of `ST_partstat' drivers/scsi/st.o(.bss+0x0): first defined here Signed-off-by: Adrian Bunk Cc: James Bottomley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] x86 current_stack_pointer warning fix With newer gcc's: include/asm/thread_info.h:95: warning: unused variable `current_stack_pointer' Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64 iSeries: fix viodasd remove This patch just makes sure that we do not dereference a viodasd gendisk pointer after it has been freed. Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: move emulate_step to arch/ppc64/lib This patch moves the emulate_step function, which is used in xmon's single-stepping code, out of xmon.c and into arch/ppc64/lib/sstep.c, so that kprobes can use it too. Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] ppc64: Make pci_alloc_consistent() conform to API docs Documentation/DMA-mapping.txt says that pci_alloc_consistent() needs to return a mapping that is aligned by the closest larger order of two as the allocation. We're currently breaking this with our iommu code. To fix this, add align_order arguments to the relevant functions and pass it down. Specifying align_order of 0 gives same behaviour as previous. Signed-off-by: Olof Johansson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] x86_64: only single-step into signal handlers if the tracer asked for it Port of the i386 patch of the same name. Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] mmtimer driver update - reduce processing in timer interrupt through the use of a tasklet - fix various race conditions - use the correct interrupt vector for the SN2 RTC Signed-off-by: Dimitri Sivanich Signed-off-by: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: Kconfig.debug support This patch updates Kconfig and add Kconfig.debug for m32r. - arch/m32r/Kconfig: Move "Kernel hacking" menu to Kconfig.debug. - arch/m32r/Kconfig.debug: Newly added. - lib/Kconfig.debug: Add m32r arch. Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: Fix a boot hang of UP kernel This patch fixes a hanging up at boot time of 2.6.10-rc2 m32r UP kernel. It was due to a lack of update_process_times() in time.c. Such a boot hang was caused only in UP systems, because update_process_times() had been executed correctly in smp_local_timer_interrupt() of arch/m32r/kernel/smp.c for SMP. arch/m32r/kernel/time.c: - UP: Fix do_timer_interrupt() to use update_process_times(). - UP: Move profile_tick() into do_timer_interrupt(). - Change __inline__ to inline. Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] m32r: make zImage a default build target Here is a patch to update arch/m32r/Makefile for m32r. - Make zImage a default build target - Add zImage to targets marked with [*]. [PATCH] m32r: io_xxxxx.c cleanups cleanups for arch/m32r/kernel/io*.c. - Fix ugly indentation. - Change __inline__ to inline. - Remove RCS ID strings. Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] M68k: Update defconfigs for 2.6.10-rc1 M68k: Update defconfigs Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] M68k: Add 3 missing syscalls M68k: Add 3 missing syscalls (up to 2.6.10-rc1) Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] 68851 MMU: Fix harmless typo in the MMU configuration code. 68851 MMU: Fix harmless (CPU_68020 == MMU_68851 anyway) typo in the MMU configuration code. Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Sun-3: Fix link error Sun-3: Fix link error (we forgot to update vmlinux-sun3.lds during last update of vmlinux-std.lds) Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] fm2fb: Update Steffen A. Mork's email address fm2fb: Update Steffen A. Mork's email address Signed-off-by: Steffen A. Mork Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] M68k I/O: Move HP300 I/O macros close to other I/O macros again M68k I/O: Move HP300 I/O macros close to other I/O macros again (after merge error in 2.6.10-rc2) Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] M68k: Update defconfigs for 2.6.10-rc2 M68k: Update defconfigs for 2.6.10-rc2 Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: update some copyrights Update/add some copyright notices. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uml: partial KBUILD_OUTPUT fix Start fixing KBUILD_OUTPUT support for UML. These changes are trivial and no-ops when this feature is not enabled - the "hard part" of this support is under discussion because it's hard to do properly (UML uses both shipped and build-generated headers from a lot of different directories, and for good reasons). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] s390: remove zfcp hba api callbacks From: Heiko Carstens Remove the disputed hba api event callback code. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] allow NFS exports of EFS filesystems Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] sonypi: return an error from sonypi_camera_command() if the camera isn't enabled The sonypi_camera_command() used to fail without returning an error code if the user fergot to enable the camera in the sonypi module (using the camera=1 module parameter). This caused the meye driver to apparently load correctly but miserably fail later, when trying to access the camera for getting some data out of it. This patch adds an error code to sonypi_camera_command() and makes the meye driver check for it in the PCI probe routine. If the function fails, a message is printed in the kernel logs reminding the user it should better RTFM. The patch also removes some sonypi_camera_command() commands (those supposed to return the current camera settings) which are unreliable. The meye driver does not use them anyway. Signed-off-by: Stelian Pop Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] uninline do_trap(), remove get_cr2() Uninlining do_trap() saves 544 bytes in traps.o. get_cr2() seems to be unused, remove it. Signed-off-by: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Switch therm_adt746x to new module_param This patch replaces MODULE_PARM to module_param for adt746x. Signed-off-by: Colin Leroy Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] unlocked access to task->comm Looking at get_task_comm patch: http://linus.bkbits.net:8080/linux-2.5/patch@1.1803.144.3 There is one other place where task->comm is accessed outside current. There are two issues. The code is trying to copy to temp space without task_lock. It is not using temp space for actual user copy. Cc: Ralf Baechle Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Permit LOG_SENSE and LOG_SELECT in SG_IO command table This patch adds LOG_SENSE as a read-ok command. cdrecord-prodvd uses this. I also added LOG_SELECT as write-ok as this seems to fit in as well. Signed-off-by: Daniel Drake Acked-by: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] knfsd: svcrpc: fqdn length fix Problem identified by Jan Kasprzak. Limit on domainname_max (currently 50) is too small. Just use the beginning of input buffer as scratch space for it, and save a little stack space while we're at it. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] documentation: nohighio The kernel parameter "nohighio" seems to be gone in the code, but the parameter is still left in the documentation. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] documentation: nmi_watchdog.txt update Acked-by: Maciej W. Rozycki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] remove outdated OSS Changelogs There's not much value in shipping Changelogs that weren't updated since at least 2.4.0. Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] linux/mount.h: add atomic.h and spinlock.h #includes uses atomic_t and spinlock_t, but doesn't include either or , which means that any users of have to include them. This patch adds the necessary #includes to avoid this. Signed-off-by: Roland Dreier Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] smbfs: Bug #3758 - Broken symlinks on smbfs The patches attached here fixes the BUG #3758 - "Broken symlinks on smbfs with 2.6.10-rc[12]". There are two patches, one to be applied over version 2.6.9 and the other over 2.6.10-rc2. The old utilities (and the old driver) uses uid=0, gid=0, dmask=0 and fmask=0 to flag the lack of this parameters on the mount command line. When the user do not specify the uid, gid, fmask or dmask, the current driver will assign gid=root, gid=root, dmask=755, fmask=755. This behavior is similar to the old 2.x samba versions. To make the driver see the permissions and ownership assigned on the server, "smbmount" and "smbmnt" utilities must be patched. The patches is already available on the attached patches at Bug #3330 or Samba Bug #999. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Minor fix of inequalities in the quota code Change inequalities for deciding when a user has cleaned up enough space to be below softlimit to match the ones for deciding when the softlimit has been exceeded. Signed-off-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] Add missing DQUOT_OFF Attached patch adds missing DQUOT_OFF to the umount path of the root filesystem (it is only remounted read-only and so the usual path with DQUOT_OFF is not taken). Signed-off-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] akpm has moved And I didn't feel a thing. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [PATCH] x86_64: fix vsyscalls Fix incorrect alignment in the vsyscall variables that caused vsyscalls to be completely broken. This change should decrease system time during TPC-* tests considerably. Clean up the vmlinux.lds to make it easier readable Do some cleanups in the vsyscall code. Align cacheline_aligned correctly on 128 byte cacheline systems. Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds [PATCH] x86_64: fix interrupt routing with nosmp Fix interaction between nosmp and pcibios_fixup_irqs(). Originally from Oleg Nesterov via i386 When we boot with nosmp we dont have all the mptable info, so IO_APIC_get_PCI_irq_vector() doesnt work and devices just end up getting a wrong interrupt. Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds [PATCH] x86_64: fix early oops printing Fix early oopses Without this very early oopses crash recursively while trying to figure out the current CPU number. Based on a patch by James Cleverdon. Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds [PATCH] x86_64: increase timer fallback threshold Increase time fallback threshold Otherwise ACPI triggers it too often. Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds [PATCH] fastcall fixes for x86 smp interrupts I cross-checked the functions called by the BUILD_INTERRUPT macros and came up with this patch. Even though some of these functions currently take no args I made them all consistent. Some functions in the Voyager code that are not directly called from asm code become static. Signed-off-by: Brian Gerst Signed-off-by: Linus Torvalds [AF_UNIX]: Don't lose ECONNRESET in unix_seqpacket_sendmsg() The fix for SELinux w/SOCK_SEQPACKET had an error, noted by Alan Cox. This fixes it. Signed-off-by: James Morris Signed-off-by: David S. Miller [PCMCIA] replace schedule_timeout() with msleep() Use msleep() instead of schedule_timeout() to guarantee the task delays as expected. Signed-off-by: Nishanth Aravamudan Signed-off-by: Maximilian Attems Signed-off-by: Russell King [PCMCIA] make cardbus_type static Patch from Adrian Bunk Since there's no user outside this file, cardbus_type in drivers/pcmcia/yenta_socket.c can be made static. Signed-off-by: Adrian Bunk Signed-off-by: Russell King [PCMCIA] Exclude uneeded code when ! CONFIG_PROC_FS Patch from Michal Rokos proc_pccard is declared twice, so remove the first declaration such that it is only in ifdef CONFIG_PROC_FS block. Signed-off-by: Russell King [ARM] Sparse fixes Clean up sparse warnings in rtctime.c and netwinder-hw.c [ARM] Add missing ecard_resource_flags() macro. [ARM] Fix VFP NaN flag handling. The special non-signalling NaN flag was escaping and setting some FPSCR exception enable flags. Ensure that this flag is masked out before we update the FPSCR. [ARM] kill unused call_irq() Patch from Christoph Hellwig These routine in arm is unused (in fact not even compiled). Instead of converting it to local_softirq_pending I'd suggest just removing it as below as it's been there totally unused for a long time. [ARM PATCH] 2267/1: don't fiddle with GPDR/GAFR directly Patch from Nicolas Pitre This is preemption unsafe and rather not pretty. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King [PATCH] sparc io.h annotations and fixes * annotations for sparc io.h * missing memcpy_toio()/memcpy_fromio() added Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] pcilynx iomem annotations Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] tgafb iomem annotations Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] iphase iomem annotations partially annotated Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] 64bit portability fixes assorted 64bit issues - extraction of number passed in pointer, use of %zd for size_t, ssize_t instead of int as return value of ->read(), %p for pointers instead of casting to int and using %x. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] ad1889 annotations and fixes fixed leaks on failure exits in ->probe() [PATCH] drivers/media annotations trivial iomem annotations Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] (1/4) eicon iomem annotations and fixes * adds new helpers ({GET,PUT}_{WORD,DWORD}) parallel to READ_WORD et.al., but used on normal memory instead of iomem. Instances of READ_WORD and friends that are applied to vmalloc'ed memory and local variables switched to new helpers. Signed-off-by: Armin Schindler Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] (2/4) eicon iomem annotations and fixes * added __iomem to declarations of iomem pointers Signed-off-by: Armin Schindler Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] (3/4) eicon iomem annotations and fixes * new helpers - READ_BYTE() and WRITE_BYTE(). On Linux they are needed, since direct access to iomem pointers is not allowed (and does not work on some architectures). * memcpy() to/from iomem is not allowed on Linux (same story). Proper primitives are memcpy_toio() and memcpy_fromio(). Several places misused memcpy(); switched to memcpy_{to,from}io(). Signed-off-by: Armin Schindler Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] (4/4) eicon iomem annotations and fixes * s_4bri.c has a nasty typo - bogus & in qBri_cpu_trapped(). We have void *base; followed by regs[0] = READ_DWORD((&base + offset) + 0x70); which is *NOT* what is meant there - instead of access to memory pointed to by base we get access to _stack_ at some offset from the place where the local variable base lives. Signed-off-by: Armin Schindler Signed-off-by: Al Viro Signed-off-by: Linus Torvalds [PATCH] x86_64: dupli