- 20 Oct, 2004 40 commits
-
-
Oleg Nesterov authored
{set,clear}_child_tid initialized in copy_process() right after return from copy_thread(). These vars are not used in cleanup path if copy_thread() fails. grep -r _child_tid arch/ shows only ia64/kernel/asm-offsets.c, so i blindly patched non i386 archs too. Signed-off-by:Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Venkatesh Pallipadi authored
phys_proc_id gets initialized only when (smp_num_siblings > 1). But gets printed even when (smp_num_siblings == 1). As a result we print incorrect physical processor id in /proc/cpuinfo, when HT is disabled. Signed-off-by:
: "Venkatesh Pallipadi" <venkatesh.pallipadi@intel.com> Signed-off-by:
-
Christoph Hellwig authored
<linux/irq.h> is internals for the generic irq handler implementation, which is used on most but not all platforms. Signed-off-by:
-
Oleg Nesterov authored
Now there is no point in calling costly find_pid(type) if __detach_pid(type) returned non zero value. Acked-By:
Kirill Korotaev <dev@sw.ru> Signed-off-by:
-
Oleg Nesterov authored
Kirill's kernel/pid.c rework broke optimization logic in detach_pid(). Non zero return from __detach_pid() was used to indicate, that this pid can probably be freed. Current version always (modulo idle threads) return non zero value, thus resulting in unneccesary pid_hash scanning. Also, uninlining __detach_pid() reduces pid.o text size from 2492 to 1600 bytes. Acked-By:
-
Christoph Lameter authored
* Add CLOCK_SGI_CYCLE provided by drivers/char/mmtimer Signed-off-by:
-
Christoph Lameter authored
POSIX clocks are to be implemented in the following way according to V3 of the Single Unix Specification: 1. CLOCK_PROCESS_CPUTIME_ID Implementations shall also support the special clockid_t value CLOCK_PROCESS_CPUTIME_ID, which represents the CPU-time clock of the calling process when invoking one of the clock_*() or timer_*() functions. For these clock IDs, the values returned by clock_gettime() and specified by clock_settime() represent the amount of execution time of the process associated with the clock. 2. CLOCK_THREAD_CPUTIME_ID Implementations shall also support the special clockid_t value CLOCK_THREAD_CPUTIME_ID, which represents the CPU-time clock of the calling thread when invoking one of the clock_*() or timer_*() functions. For these clock IDs, the values returned by clock_gettime() and specified by clock_settime() shall represent the amount of execution time of the thread associated with the clock. These times mentioned are CPU processing times and not the time that has passed since the startup of a process. Glibc currently provides its own implementation of these two clocks which is designed to return the time that passed since the startup of a process or a thread. Moreover Glibc's clocks are bound to CPU timers which is problematic when the frequency of the clock changes or the process is moved to a different processor whose cpu timer may not be fully synchronized to the cpu timer of the current CPU. This patchset results in a both clocks working reliably. The patch also implements the access to other the thread and process clocks of linux processes by using negative clockid's: 1. For CLOCK_PROCESS_CPUTIME_ID: -pid 2. For CLOCK_THREAD_CPUTIME_ID: -(pid + PID_MAX_LIMIT) This allows clock_getcpuclockid(pid) to return -pid and pthread_getcpuiclock(pid) to return -(pid + PID_MAX_LIMIT) to allow access to the corresponding clocks. Todo: - The timer API to generate events by a non tick based timer is not usable in its current state. The posix timer API seems to be only useful at this point to define clock_get/set. Need to revise this. - Implement timed interrupts in mmtimer after API is revised. The mmtimer patch is unchanged from V6 and stays as is in 2.6.9-rc3-mm2. But I expect to update the driver as soon as the interface to setup hardware timer interrupts is usable. Single Thread Testing CLOCK_THREAD_CPUTIME_ID= 0.494140878 resolution= 0.000976563 CLOCK_PROCESS_CPUTIME_ID= 0.494140878 resolution= 0.000976563 Multi Thread Testing Starting Thread: 0 1 2 3 4 5 6 7 8 9 Joining Thread: 0 1 2 3 4 5 6 7 8 9 0 Cycles= 0 Thread= 0.000000000ns Process= 0.495117441ns 1 Cycles=1000000 Thread= 0.140625072ns Process= 2.523438792ns 2 Cycles=2000000 Thread= 0.966797370ns Process= 8.512699671ns 3 Cycles=3000000 Thread= 0.806641038ns Process= 7.561527309ns 4 Cycles=4000000 Thread= 1.865235330ns Process= 12.891608163ns 5 Cycles=5000000 Thread= 1.604493009ns Process= 11.528326215ns 6 Cycles=6000000 Thread= 2.086915131ns Process= 13.500983475ns 7 Cycles=7000000 Thread= 2.245118337ns Process= 13.947272766ns 8 Cycles=8000000 Thread= 1.604493009ns Process= 12.252935961ns 9 Cycles=9000000 Thread= 2.160157356ns Process= 13.977546219ns Clock status at the end of the timer tests: Gettimeofday() = 1097084999.489938000 CLOCK_REALTIME= 1097084999.490116229 resolution= 0.000000040 CLOCK_MONOTONIC= 177.071675109 resolution= 0.000000040 CLOCK_PROCESS_CPUTIME_ID= 13.978522782 resolution= 0.000976563 CLOCK_THREAD_CPUTIME_ID= 0.497070567 resolution= 0.000976563 CLOCK_SGI_CYCLE= 229.967982280 resolution= 0.000000040 PROCESS clock of 1 (init)= 4.833986850 resolution= 0.000976563 THREAD clock of 1 (init)= 0.009765630 resolution= 0.000976563 Signed-off-by: -
Frank Hirtz authored
The following patch will have the committed memory limit (per the current overcommit ratio) and the amount of memory remaining under this limit displayed in meminfo. It's presently somewhat difficult to use the strict memory overcommit settings as it's somewhat difficult to determine the amount of memory remaining under the cap. This patch would make using strict overcommit a good bit simpler. Signed-off-by:
-
Stephen Hemminger authored
A couple chrdev routines take a constant string and should be declared with const char *. Signed-off-by:
-
Michael A. Halcrow authored
This patch includes documentation on using the BSD Secure Levels LSM. Signed-off-by:
Michael A. Halcrow <mahalcro@us.ibm.com> Signed-off-by:
-
Michael A. Halcrow authored
This patch modifies Kconfig and Makefile to support building the BSD Secure Levels LSM, in addition to the module itself. Signed-off-by:
-
Michael A. Halcrow authored
I have received positive feedback from various individuals who have applied my BSD Secure Levels LSM patch, and so at this point I am submitting it to you with a request to merge it in. Nothing has changed in this patch since when I last posted it to the LKML, so I am not re-sending it there. This first patch adds hooks to catch attempts to set the system clock back. Signed-off-by:
-
Roland McGrath authored
I don't know why struct k_itimer was ever declared in sched.h; perhaps at one time it was referenced by something else there. There is no need for it now. This patch moves the struct where it belongs, in linux/posix-timers.h. It has zero effect on anything except keeping the source easier to read. Signed-off-by:
-
Jan-Benedict Glaw authored
Signed-off-by:
-
Ingo Molnar authored
The per-cpu disk stats are being updated in a non-preempt-safe manner in a couple of places. The patch introduces introduces preempt and non-preempt versions of the statistics code and updates the block code to use the appropriate ones. Signed-off-by:
-
Zwane Mwaikambo authored
The 'noapic' kernel parameter only disables IOAPIC use and not all the APICs (which would include local APICs) in the system. Signed-off-by:
Zwane Mwaikambo <zwane@fsmlabs.com> Signed-off-by:
-
William Lee Irwin III authored
This patch converts all users of remap_page_range() under sound/ to use remap_pfn_range(), with the exception of maestro3 changelogs, which are likely expected to be preserved intact apart from additions (as most changelogs are), regardless of API changes. Signed-off-by:
-
William Lee Irwin III authored
This patch converts uses of remap_page_range() via io_remap_page_range() in include/asm-*/ to use remap_pfn_range(). io_remap_page_range() has a similar physical address overflow issue that needs to be addressed later. Signed-off-by:
-
William Lee Irwin III authored
This patch converts all callers of remap_page_range() under arch/ and net/ to use remap_pfn_range() instead. Signed-off-by:
-
William Lee Irwin III authored
[PATCH] vm: convert references to remap_page_range() under arch/ and Documentation/ to remap_pfn_range() This patch converts all callers of remap_page_range() under arch/ and all references in Documentation/ to use remap_pfn_range(). Signed-off-by:
-
William Lee Irwin III authored
This patch introduces remap_pfn_range(), destined to replace remap_page_range(), to which all callers of remap_page_range() are converted in the sequel. Signed-off-by:
-
Neil Brown authored
for OPEN with O_TRUNC, if the truncate fails, the open fails. for nfs4_open_upgrade, this means undo the get_write_access. for new OPENs, this means release the newly created stateid. Signed-off-by:
Andy Adamson <andros@citi.umich.edu> Signed-off-by:
J. Bruce Fields <bfields@citi.umich.edu> Signed-off-by:
Neil Brown <neilb@cse.unsw.edu.au> Signed-off-by:
-
Neil Brown authored
Thanks to Al Viro for noticing that putrootfh could return either a linux or an nfs error. Signed-off-by:
-
Neil Brown authored
Comments, dprintk cleanup Signed-off-by:
-
Neil Brown authored
Move most of the code in the new_open case of nfsd4_process_open2 to nfs4_new_open. Signed-off-by:
Andy Adamson <andros@umich.edu> Signed-off-by:
-
Neil Brown authored
Move most of the code for the open_upgrade case into a separate nfs4_upgrade_open. Signed-off-by:
-
Neil Brown authored
Reorganize "if" in nfsd4_process_open2 to make test clearer; there was no real reason to test for the negative here. Signed-off-by:
-
Neil Brown authored
Move seqid decrement on reclaim to separate function Signed-off-by:
-
Neil Brown authored
We should set the OPEN_RESULT_LOCKTYPE_POSIX flag in the return to open, to indicate that we support locking. Signed-off-by:
-
Neil Brown authored
nfsd4_process_open2 has become a bit long and contorted. The following patches break nfsd4_process_open2 into smaller functions and add comments to describe logic flow, in preparation for delegation state. We begin by pulling out the code that searches for conflicting open owners into a separate function. Signed-off-by:
-
Neil Brown authored
When we upgrade an open, we keep the same stateid and bump only the seqid. Signed-off-by:
-
Neil Brown authored
We take a reference on the stateowner, and copy the clientid, instead of just hoping no one destroys the stateowner before we reference it in nfsd4_encode_lock_denied. Signed-off-by:
-
Neil Brown authored
With the addition of a reference count, we no longer need the code that conditionally extended the nfs4 state lock over the xdr encode stage. Signed-off-by:
-
Neil Brown authored
Take a reference to preserve the stateowner through the xdr replay code, and simplify nfsd4_proc_compound a little. Signed-off-by:
-
Neil Brown authored
Reference-counting the nfsd4 stateowner structs will let us fix a race and simplify some of the xdr code a bit, and may also help us make the nfsd4 locking a little more fine-grained in the future. Signed-off-by:
-
Neil Brown authored
Make sure getxattr inode op is non-NULL before calling it. Also, security hook should probably be called before calling the getxattr op the first time. Signed-off-by:
-
Neil Brown authored
The ->getxattr op doesn't take the i_sem (see Documentation/filesystems/Locking) Signed-off-by:
-
Neil Brown authored
Separate out the little bit of logic in fh_verify that checks the i_mode into a separate function. This is nicely encapsulated, fh_verify() is overly long anyway, and this change helps make a subsequent nfs4 named attribute change clearer. Signed-off-by:
-
Neil Brown authored
Bug Fix: Non NFSD conflicting byte-range locks were causing an Oops. Encode a zero length owner and zero clientid for non NFSD conflicting locks in the lock_denied response. Signed-off-by:
-
Andrew Morton authored
Kconfig barfs if it doesn't have leading spaces in the help. Signed-off-by:
-
