- 08 May, 2007 40 commits
-
-
Alexey Dobriyan authored
Eternal quest to make while true; do cat /proc/fs/xfs/stat >/dev/null 2>/dev/null; done while true; do find /proc -type f 2>/dev/null | xargs cat >/dev/null 2>/dev/null; done while true; do modprobe xfs; rmmod xfs; done work reliably continues and now kernel oopses in the following way: BUG: unable to handle ... at virtual address 6b6b6b6b EIP is at badness process: cat proc_oom_score proc_info_read sys_fstat64 vfs_read proc_info_read sys_read Failing code is prefetch hidden in list_for_each_entry() in badness(). badness() is reachable from two points. One is proc_oom_score, another is out_of_memory() => select_bad_process() => badness(). Second path grabs tasklist_lock, while first doesn't. Signed-off-by:
Alexey Dobriyan <adobriyan@sw.ru> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Nick Piggin authored
LTP test sigaction_16_24 fails, because it expects sem_wait to be restarted if SA_RESTART is set. sem_wait is implemented with futex_wait, that currently doesn't support being restarted. Ulrich confirms that the call should be restartable. Implement a restart_block method to handle the relative timeout, and allow restarts. Signed-off-by:
Nick Piggin <npiggin@suse.de> Cc: Ulrich Drepper <drepper@redhat.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Roland McGrath <roland@redhat.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Acked-by:
Ingo Molnar <mingo@elte.hu> Signed-off-by:
-
Rusty Russell authored
lguest uses the convenient futex infrastructure for inter-domain I/O, so expose get_futex_key, get_key_refs (renamed get_futex_key_refs) and drop_key_refs (renamed drop_futex_key_refs). Also means we need to expose the union that these use. No code changes. Signed-off-by:
Rusty Russell <rusty@rustcorp.com.au> Cc: Andi Kleen <ak@suse.de> Signed-off-by:
-
Dale Farnsworth authored
Signed-off-by:
Dale Farnsworth.org <dale@farnsworth.org> Cc: Alessandro Zummo <a.zummo@towertech.it> Cc: David Brownell <david-b@pacbell.net> Cc: Jean Delvare <khali@linux-fr.org> Signed-off-by:
-
Nobuhiro Iwamatsu authored
Add an RTC driver for Ricoh RS5C313 RTC chip. [akpm@linux-foundation.org: Zillions of coding-style fixes] [akpm@linux-foundation.org: build fixes] Signed-off-by:
Nobuhiro Iwamatsu <iwamatsu@nigauri.org> Cc: Alessandro Zummo <a.zummo@towertech.it> Cc: David Brownell <david-b@pacbell.net> Signed-off-by:
-
Nick Piggin authored
Fix units mismatch (jiffies vs msecs) in as-iosched.c, spotted by Xiaoning Ding <dingxn@cse.ohio-state.edu>. Signed-off-by:
-
Len Sorensen authored
The jsm driver doesn't currently use the uart_handle_*_change helper functions, which are the obvious place for things like linuxpps to tie into (which it now does of course), and as a result the jsm driver can not be used with linuxpps and anything else that ties into the serial_core helper functions. This patch adds calls to these helper functions whenever the value they manage changes. That actual storage of the state is not modified since the jsm driver caches the current settings (The 8250 driver reads them everytime a user asks for the state), and only updates them whenever they change. Signed-off-by:
Len Sorensen <lsorense@csclub.uwaterloo.ca> Cc: Scott H Kilau <Scott_Kilau@digi.com> Cc: Wendy Xiong <wendyx@us.ibm.com> Signed-off-by:
-
Len Sorensen authored
The jsm driver fails when you try to use the TIOCSSERIAL ioctl. The reason is that the driver never sets uart_port.uartclk, causing the data received using TIOCGSERIAL to not match the internal state of the driver. This patch fixes this problem by settings the uartclk to the value used by the serial_core (16 times the baud base). Signed-off-by:
-
Klaus Kudielka authored
Switch from private uclong, etc over to standard types. Signed-off-by:
-
Klaus Kudielka authored
At least on x86_64 the present cyclades.h is broken due to the wrong size of uclong. This affects, of course, both the kernel and the user-level utilities. The symptom is that cyzload refuses to load the firmware. I also managed to freeze the machine when unloading the module. The patch below fixes this in an architecture-independent way. I have tested it with 2.6.19 and the driver works fine again with a Cyclades-Z on an Athlon 64 X2. [akpm@linux-foundation.org: fix warnings] Signed-off-by:
-
Ananth N Mavinakayanahalli authored
Kprobes doesn't scribble the kprobe.symbol_name field. Its only set by the module when registering the probe. Modules that exercise good hygiene using the "const" qualifier will see warnings... warning: assignment discards qualifiers from pointer target type Make struct kprobe.symbol_name const char * Signed-off-by:
Ananth N Mavinakayanahalli <ananth@in.ibm.com> Signed-off-by:
Jim Keniston <jkenisto@us.ibm.com> Signed-off-by:
-
Alan Cox authored
Adds the needed TCGETS2/TCSETS2 ioctl calls, structures, defines and the like. Tested against the test suite and passes. Other platforms should need roughly the same change. Signed-off-by:
Alan Cox <alan@redhat.com> Signed-off-by:
-
Eric Dumazet authored
1) Introduces a new method in 'struct dentry_operations'. This method called d_dname() might be called from d_path() to build a pathname for special filesystems. It is called without locks. Future patches (if we succeed in having one common dentry for all pipes/sockets) may need to change prototype of this method, but we now use : char *d_dname(struct dentry *dentry, char *buffer, int buflen); 2) Adds a dynamic_dname() helper function that eases d_dname() implementations 3) Defines d_dname method for sockets : No more sprintf() at socket creation. This is delayed up to the moment someone does an access to /proc/pid/fd/... 4) Defines d_dname method for pipes : No more sprintf() at pipe creation. This is delayed up to the moment someone does an access to /proc/pid/fd/... A benchmark consisting of 1.000.000 calls to pipe()/close()/close() gives a *nice* speedup on my Pentium(M) 1.6 Ghz : 3.090 s instead of 3.450 s Signed-off-by:
Eric Dumazet <dada1@cosmosbay.com> Acked-by:
Christoph Hellwig <hch@infradead.org> Acked-by:
-
Miklos Szeredi authored
Add support for finding out the current file position, open flags and possibly other info in the future. These new entries are added: /proc/PID/fdinfo/FD /proc/PID/task/TID/fdinfo/FD For each fd the information is provided in the following format: pos: 1234 flags: 0100002 [bunk@stusta.de: make struct proc_fdinfo_file_operations static] Signed-off-by:
Miklos Szeredi <mszeredi@suse.cz> Cc: Alexey Dobriyan <adobriyan@sw.ru> Signed-off-by:
Adrian Bunk <bunk@stusta.de> Signed-off-by:
-
Eric Dumazet authored
Change the order of fields of struct pid_entry (file fs/proc/base.c) in order to avoid a hole on 64bit archs. (8 bytes saved per object) Also change all pid_entry arrays to be const qualified, to make clear they must not be modified. Before (on x86_64) : # size fs/proc/base.o text data bss dec hex filename 15549 2192 0 17741 454d fs/proc/base.o After : # size fs/proc/base.o text data bss dec hex filename 17229 176 0 17405 43fd fs/proc/base.o Thats 336 bytes saved on kernel size on x86_64 Signed-off-by:
"Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by:
-
Amit Choudhary authored
Signed-off-by:
Amit Choudhary <amit2030@gmail.com> Cc: Paul Fulghum <paulkf@microgate.com> Signed-off-by:
-
Robert P. J. Day authored
Signed-off-by:
Robert P. J. Day <rpjday@mindspring.com> Signed-off-by:
-
Robert P. J. Day authored
Signed-off-by:
-
Kees Cook authored
The /proc/pid/ "maps", "smaps", and "numa_maps" files contain sensitive information about the memory location and usage of processes. Issues: - maps should not be world-readable, especially if programs expect any kind of ASLR protection from local attackers. - maps cannot just be 0400 because "-D_FORTIFY_SOURCE=2 -O2" makes glibc check the maps when %n is in a *printf call, and a setuid(getuid()) process wouldn't be able to read its own maps file. (For reference see http://lkml.org/lkml/2006/1/22/150) - a system-wide toggle is needed to allow prior behavior in the case of non-root applications that depend on access to the maps contents. This change implements a check using "ptrace_may_attach" before allowing access to read the maps contents. To control this protection, the new knob /proc/sys/kernel/maps_protect has been added, with corresponding updates to the procfs documentation. [akpm@linux-foundation.org: build fixes] [akpm@linux-foundation.org: New sysctl numbers are old hat] Signed-off-by:
Kees Cook <kees@outflux.net> Cc: Arjan van de Ven <arjan@infradead.org> Signed-off-by:
-
Andrew Morton authored
WARNING: vmlinux - Section mismatch: reference to .init.text:eisa_root_register from .text between 'virtual_eisa_root_init' (at offset 0xc026b80f) and 'cpufreq_debug_disable_ratelimit' Cc: Dave Jones <davej@codemonkey.org.uk> Signed-off-by:
-
Robert P. J. Day authored
It misspelled "MODVERSIONS" preprocessor variable with "CONFIG_MODVERSIONS". Just kill it all. Signed-off-by:
-
Dmitriy Monakhov authored
This patch kills the "ignoring return value of 'device_create_file'" warning message. Signed-off-by:
Monakhov Dmitriy <dmonakhov@openvz.org> Signed-off-by:
-
Surya authored
Cleaning up of pci_find_device in drivers/telephony/ixj.c. Signed-off-by:
Surya Prabhakar <surya.prabhakar@wipro.com> Signed-off-by:
-
Alex Williamson authored
tAdd adds support for devices living in MMIO space to the Infineon TPM driver. These can be found on some of the newer HP ia64 systems. Signed-off-by:
Alex Williamson <alex.williamson@hp.com> Cc: Kylene Jo Hall <kjhall@us.ibm.com> Acked-by:
Marcel Selhorst <tpm@selhorst.net> Signed-off-by:
-
Christoph Hellwig authored
We don't have a routine called namei() anymore since at least 2.3.x, and the comment is just totally out of sync with the current lookup logic. Signed-off-by:
Christoph Hellwig <hch@lst.de> Signed-off-by:
-
Christoph Hellwig authored
inode->i_sb is always set, not need to check for it. Signed-off-by:
-
Alexey Dobriyan authored
WARN_ON(de && de->deleted); is sooo unreliable. Why? proc_lookup remove_proc_entry =========== ================= lock_kernel(); spin_lock(&proc_subdir_lock); [find proc entry] spin_unlock(&proc_subdir_lock); spin_lock(&proc_subdir_lock); [find proc entry] proc_get_inode ============== WARN_ON(de && de->deleted); ... if (!atomic_read(&de->count)) free_proc_entry(de); else de->deleted = 1; So, if you have some strange oops [1], and doesn't see this WARN_ON it means nothing. [1] try_module_get() of module which doesn't exist, two lines below should suffice, or not? Signed-off-by: -
Darrick J. Wong authored
Fix the following race: proc_readdir remove_proc_entry ============ ================= spin_lock(&proc_subdir_lock); [choose PDE to start filldir from] spin_unlock(&proc_subdir_lock); spin_lock(&proc_subdir_lock); [find PDE] [free PDE, refcount is 0] spin_unlock(&proc_subdir_lock); /* boom */ if (filldir(dirent, de->name, ... [de_put on error path --adobriyan] Signed-off-by:
Darrick J. Wong <djwong@us.ibm.com> Signed-off-by:
-
Alexey Dobriyan authored
proc_lookup remove_proc_entry =========== ================= lock_kernel(); spin_lock(&proc_subdir_lock); [find PDE with refcount 0] spin_unlock(&proc_subdir_lock); spin_lock(&proc_subdir_lock); [find PDE with refcount 0] [check refcount and free PDE] spin_unlock(&proc_subdir_lock); proc_get_inode: de_get(de); /* boom */ Signed-off-by:
Alexey Dobriyan <adobriyan@openvz.org> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by:
-
Miklos Szeredi authored
There's a slight problem with filesystem type representation in fuse based filesystems. From the kernel's view, there are just two filesystem types: fuse and fuseblk. From the user's view there are lots of different filesystem types. The user is not even much concerned if the filesystem is fuse based or not. So there's a conflict of interest in how this should be represented in fstab, mtab and /proc/mounts. The current scheme is to encode the real filesystem type in the mount source. So an sshfs mount looks like this: sshfs#user@server:/ /mnt/server fuse rw,nosuid,nodev,... This url-ish syntax works OK for sshfs and similar filesystems. However for block device based filesystems (ntfs-3g, zfs) it doesn't work, since the kernel expects the mount source to be a real device name. A possibly better scheme would be to encode the real type in the type field as "type.subtype". So fuse mounts would look like this: /dev/hda1 /mnt/windows fuseblk.ntfs-3g rw,... user@server:/ /mnt/server fuse.sshfs rw,nosuid,nodev,... This patch adds the necessary code to the kernel so that this can be correctly displayed in /proc/mounts. Signed-off-by:
-
Jean Delvare authored
strlcpy already accounts for the trailing zero in its length computation, so there is no need to substract one to the buffer size. Signed-off-by:
Jean Delvare <khali@linux-fr.org> Signed-off-by:
-
Davide Libenzi authored
Epoll is doing multiple passes over the ready set at the moment, because of the constraints over the f_op->poll() call. Looking at the code again, I noticed that we already hold the epoll semaphore in read, and this (together with other locking conditions that hold while doing an epoll_wait()) can lead to a smarter way [1] to "ship" events to userspace (in a single pass). This is a stress application that can be used to test the new code. It spwans multiple thread and call epoll_wait() and epoll_ctl() from many threads. Stress tested on my dual Opteron 254 w/out any problems. http://www.xmailserver.org/totalmess.c This is not a benchmark, just something that tries to stress and exploit possible problems with the new code. Also, I made a stupid micro-benchmark: http://www.xmailserver.org/epwbench.c [1] Considering that epoll must be thread-safe, there are five ways we can be hit during an epoll_wait() transfer loop (ep_send_events()): 1) The epoll fd going away and calling ep_free This just can't happen, since we did an fget() in sys_epoll_wait 2) An epoll_ctl(EPOLL_CTL_DEL) This can't happen because epoll_ctl() gets ep->sem in write, and we're holding it in read during ep_send_events() 3) An fd stored inside the epoll fd going away This can't happen because in eventpoll_release_file() we get ep->sem in write, and we're holding it in read during ep_send_events() 4) Another epoll_wait() happening on another thread They both can be inside ep_send_events() at the same time, we get (splice) the ready-list under the spinlock, so each one will get its own ready list. Note that an fd cannot be at the same time inside more than one ready list, because ep_poll_callback() will not re-queue it if it sees it already linked: if (ep_is_linked(&epi->rdllink)) goto is_linked; Another case that can happen, is two concurrent epoll_wait(), coming in with a userspace event buffer of size, say, ten. Suppose there are 50 event ready in the list. The first epoll_wait() will "steal" the whole list, while the second, seeing no events, will go to sleep. But at the end of ep_send_events() in the first epoll_wait(), we will re-inject surplus ready fds, and we will trigger the proper wake_up to the second epoll_wait(). 5) ep_poll_callback() hitting us asyncronously This is the tricky part. As I said above, the ep_is_linked() test done inside ep_poll_callback(), will guarantee us that until the item will result linked to a list, ep_poll_callback() will not try to re-queue it again (read, write data on any of its members). When we do a list_del() in ep_send_events(), the item will still satisfy the ep_is_linked() test (whatever data is written in prev/next, it'll never be its own pointer), so ep_poll_callback() will still leave us alone. It's only after the eventual smp_mb()+INIT_LIST_HEAD(&epi->rdllink) that it'll become visible to ep_poll_callback(), but at the point we're already past it. [akpm@osdl.org: 80 cols] Signed-off-by:
Davide Libenzi <davidel@xmailserver.org> Signed-off-by:
-
Adrian Bunk authored
Signed-off-by:
-
Dmitriy Monakhov authored
- ext3_dx_find_entry() exit with out setting proper error pointer - do_split() exit with out setting proper error pointer it is realy painful because many callers contain folowing code: de = do_split(handle,dir, &bh, frame, &hinfo, &retval); if (!(de)) return retval; <<< WOW retval wasn't changed by do_split(), so caller failed <<< but return SUCCESS :) - Rearrange do_split() error path. Current error path is realy ugly, all this up and down jump stuff doesn't make code easy to understand. [dmonakhov@sw.ru: fix annoying fake error messages] Signed-off-by: -
Eric Dumazet authored
The first thing done by timespec_trunc() is : if (gran <= jiffies_to_usecs(1) * 1000) This should really be a test against a constant known at compile time. Alas, it isnt. jiffies_to_usec() was unilined so C compiler emits a function call and a multiply to compute : a CONSTANT. mov $0x1,%edi mov %rbx,0xffffffffffffffe8(%rbp) mov %r12,0xfffffffffffffff0(%rbp) mov %edx,%ebx mov %rsi,0xffffffffffffffc8(%rbp) mov %rsi,%r12 callq ffffffff80232010 <jiffies_to_usecs> imul $0x3e8,%eax,%eax cmp %ebx,%eax This patch reorders kernel/time.c a bit so that jiffies_to_usecs() is defined before timespec_trunc() so that compiler now generates : cmp $0x3d0900,%edx (HZ=250 on my machine) This gives a better code (timespec_trunc() becoming a leaf function), and shorter kernel size as well. Signed-off-by:
-
David Brownell authored
PNP now initializes device dma masks, which prevents oopses when generic dma calls are made using pnp device nodes. This assumes PNP only uses ISA DMA, with 24 bit addresses; and that it's safe to init those masks for all devices (rather than finding out which devices have been assigned DMA channels, and handling only those). Signed-off-by:
David Brownell <dbrownell@users.sourceforge.net> Cc: Adam Belay <abelay@novell.com> Cc: Jaroslav Kysela <perex@suse.cz> Signed-off-by:
-
Josh Triplett authored
The corresponding rcu_torture_cleanup cannot get marked as __exit, because rcu_torture_init uses it to clean up if init fails. Signed-off-by:
Josh Triplett <josh@freedesktop.org> Acked-by:
"Paul E. McKenney" <paulmck@us.ibm.com> Signed-off-by:
-
Badari Pulavarty authored
sys_clone() and sys_unshare() both makes copies of nsproxy and its associated namespaces. But they have different code paths. This patch merges all the nsproxy and its associated namespace copy/clone handling (as much as possible). Posted on container list earlier for feedback. - Create a new nsproxy and its associated namespaces and pass it back to caller to attach it to right process. - Changed all copy_*_ns() routines to return a new copy of namespace instead of attaching it to task->nsproxy. - Moved the CAP_SYS_ADMIN checks out of copy_*_ns() routines. - Removed unnessary !ns checks from copy_*_ns() and added BUG_ON() just incase. - Get rid of all individual unshare_*_ns() routines and make use of copy_*_ns() instead. [akpm@osdl.org: cleanups, warning fix] [clg@fr.ibm.com: remove dup_namespaces() declaration] [serue@us.ibm.com: fix CONFIG_IPC_NS=n, clone(CLONE_NEWIPC) retval] [akpm@linux-foundation.org: fix build with CONFIG_SYSVIPC=n] Signed-off-by:
Badari Pulavarty <pbadari@us.ibm.com> Signed-off-by:
Serge Hallyn <serue@us.ibm.com> Cc: Cedric Le Goater <clg@fr.ibm.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: <containers@lists.osdl.org> Signed-off-by:
Cedric Le Goater <clg@fr.ibm.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by:
-
Nick Piggin authored
Petr Tesarik discovered a problem in remove_arg_zero(). He writes: When a script is loaded, load_script() replaces argv[0] with the name of the interpreter and the filename passed to the exec syscall. However, there is no guarantee that the length of the interpreter name plus the length of the filename is greater than the length of the original argv[0]. If the difference happens to cross a page boundary, setup_arg_pages() will call put_dirty_page() [aka install_arg_page()] with an address outside the VMA. Therefore, remove_arg_zero() must free all pages which would be unused after the argument is removed. So, rewrite the remove_arg_zero function without gotos, with a few comments, and with the commonly used explicit index/offset. This fixes the problem and makes it easier to understand as well. [a.p.zijlstra@chello.nl: add comment] Signed-off-by:
Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by:
-
Guy Streeter authored
The value of shmmax may be larger than will fit in the struct used by the 32bit compat version of sys_shmctl. This change mirrors what the normal sys_shmctl does when called with the old IPC_INFO command. Signed-off-by:
Guy Streeter <streeter@redhat.com> Signed-off-by:
-
