- 05 Feb, 2005 7 commits
-
-
Fabio Massimo Di Nitto authored
The patch fixes the noexec= boot option on x86_64 to actually work when other options come after it. Credits (if any ;)) should go to Matt Zimmerman and Colin Watson for spotting the problem and providing/testing the fix. Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Andi Kleen authored
This patch fixes a compile problem on x86-64 when CONFIG_PM is turned off. Signed-off-by:
Andi Kleen <ak@suse.de> Signed-off-by:
-
Matthias-Christian Ott authored
The Pentium4 models 0&1 have a longer MSR_EBC_FREQUENCY_ID register as the models 2&3, so the bit shift must be bigger. Signed-off-by:
Matthias-Christian Ott <matthias.christian@tiscali.de> Signed-off-by:
Dominik Brodowski <linux@brodo.de> Signed-off-by:
-
Hugh Dickins authored
The hugetlb_page test in do_munmap is too permissive. It checks start vma, but forgets that end vma might be different and huge though start is not: so hits unmap_hugepage_range BUG if misaligned end was given. And it's too restrictive: munmap has always succeeded on unmapped areas within its range, why should it behave differently near a hugepage vma? And the additional checks in is_aligned_hugepage_range are irrelevant here, when the hugepage vma already exists. But the function is still required (on some arches), as the default for prepare_hugepage_range - leave renaming cleanup to another occasion. Signed-off-by:
Hugh Dickins <hugh@veritas.com> Acked-by:
William Irwin <wli@holomorphy.com> Signed-off-by:
-
Andreas Gruenbacher authored
When looking for identical xattr blocks to share, we were not comparing the name_index fields. This could lead to false sharing when two xattr blocks ended up with identical attribute names and values, and the only default acls. Because acls are cached, the bug was hidden until the next reload of the affected inode. $ mkdir -m 700 a b $ setfacl -m u:bin:rwx a < acl of a goes in the mbcache $ setfacl -dm u:bin:rwx b < acl of b differs only in name_index, so a's acl is reused $ getfacl b < shows the result from the inode cache < empty inode cache (remount, etc.) $ getfacl b < shows an access acl instead of a default acl. Signed-off-by:
Andreas Gruenbacher <agruen@suse.de> Signed-off-by:
-
Hugh Dickins authored
It's time to remove truncate_complete_page's BUG_ON(page_mapped(page)): it was there to give confidence in the new vm_truncate_count mechanism. Earlier releases had no such check, and it wouldn't be at all helpful if it ever bugged up file truncation on a production system - though we don't know of any scenario in which that could happen now. Signed-off-by:
-
Hugh Dickins authored
Just before removing truncate_complete_page's BUG_ON(page_mapped(page)), thought I'd recheck on a few filesystems. The shame! Easily triggered with tmpfs: not because of recent changes, but because shmem_nopage omitted the i_size_read from Andrea's careful truncate_count/i_size_read /cachelookup/truncate_count sequence. For varying reasons, other users of shmem_getpage can't go beyond i_size, so just add it to shmem_nopage. Signed-off-by:
-
- 04 Feb, 2005 33 commits
-
-
Linus Torvalds authored
-
Paul Mackerras authored
Currently, the kernel won't compile for SMP ppc32 if preempt is enabled. This patch adds suitable read_can_lock and write_can_lock definitions. This patch also adds a real _raw_read_trylock (inline and out-of-line versions), changes the rwlock->lock field to a signed int, which is what it really was all along, and cleans up the out-of-line rwlock code in arch/ppc/lib/locks.c. It removes the debug fields from the rwlock struct because we were never using them, even with CONFIG_DEBUG_SPINLOCK set. I have compile and boot tested this with the four combinations of CONFIG_DEBUG_SPINLOCKS on and off, and CONFIG_PREEMPT on and off. Please put this patch into 2.6.11 so that 2.6.11 will work for ppc32 with CONFIG_SMP and CONFIG_PREEMPT. Signed-off-by:
Paul Mackerras <paulus@samba.org> Signed-off-by:
-
Venkatesh Pallipadi authored
This patch fixes the issue with HPET on some platforms. According to Vojtech Pavlik: The first write after writing TN_SETVAL to the config register sets the counter value, the second write sets the threshold. When you only do the first write you never set the threshold and interrupts won't be generated properly. Thanks to John Stultz and Andrew Walrond for reporting, root causing the issue and verifying this fix. Signed-off-by:
Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> Signed-off-by:
-
Stephen Rothwell authored
This patch just replaces the last usage of the vio dma mapping routines with the equivalent generic dma mapping routines. Signed-off-by:
Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by:
-
Nathan Lynch authored
Make the physical_id cpu sysfs attribute on ppc64 show -1 instead of 65535 for non-present cpus. Signed-off-by:
Nathan Lynch <nathanl@austin.ibm.com> Signed-off-by:
-
Paul Mackerras authored
This patch is from David Woodhouse <dwmw2@infradead.org>. We were pretending that every syscall returned zero. Don't do that. Signed-Off-By:
David Woodhouse <dwmw2@infradead.org> Signed-off-by:
-
Anton Blanchard authored
The 4 level pagetable code changed the exit_mmap code to rely on TASK_SIZE. On some architectures (eg ppc64 and ia64), this is a per task property and bad things can happen in certain circumstances when using it. It is possible for one task to end up "owning" an mm from another - we have seen this with the procfs code when process 1 accesses /proc/pid/cmdline of process 2 while it is exiting. Process 2 exits but does not tear its mm down. Later on process 1 finishes with the proc file and the mm gets torn down at this point. Now if process 1 was 32bit and process 2 was 64bit then we end up using a bad value for TASK_SIZE in exit_mmap. We only tear down part of the address space and leave half initialised pagetables and entries in the MMU etc. MM_VM_SIZE() was created for this purpose (and is used in the next line for tlb_finish_mmu), so use it. I moved the PGD round up of TASK_SIZE into the default MM_VM_SIZE. Signed-off-by:
Anton Blanchard <anton@samba.org> Signed-off-by:
-
David Woodhouse authored
Bad things can happen if a 32-bit process is the last user of a 64-bit mm. TASK_SIZE isn't a constant, and we can end up clearing page tables only up to the 32-bit TASK_SIZE instead of all the way. We should probably double-check every instance of TASK_SIZE or USER_PTRS_PER_PGD for this kind of problem. We should also double-check that MM_VM_SIZE() and other such things are correctly defined on all architectures. I already fixed ppc64 which let it stay as TASK_SIZE, and hence dependent on the _current_ context instead of the mm in the argument. Signed-off-by:
-
Tom Rini authored
Changeset 1.1938.196.11 broke MPC8260 PCI9 Errata workaround. This pach makes it work again. Signed-off-by:
Rune Torgersen <runet@innovsys.com> Signed-off-by:
Tom Rini <trini@kernel.crashing.org> Signed-off-by:
-
Jeff Dike authored
Forgot to use ARCH_USER_CFLAGS after defining it for x86_64. Signed-off-by:
Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Signed-off-by:
-
Jeff Dike authored
This fixes a bug which assumes that __binary_start starts on a page boundary, which isn't true when UML is configured to load into the normal executable area. Signed-off-by:
-
Jeff Dike authored
The previous ifdef to check whether to use the host's vsyscall page was buggy. This bug can cause crashes. Signed-off-by:
-
Jeff Dike authored
Initialize jiffies_64 to INITIAL_JIFFIES. Signed-off-by:
-
Bodo Stroesser authored
When a page fault occurs on an address below the stack-vma, UML tries to expand the stack. On i386 and x86_64, the failing address is compared to the current userspace stack pointer. If the failing address is below "esp-32" resp. "rsp-128", stack expansion is not allowed, and a SIGSEGV is given to the user. This patch makes UML behave like i386/x86_64. Signed-off-by:
Bodo Stroesser <bstroesser@fujitsu-siemens.com> Signed-off-by:
-
Jeff Dike authored
This adds code which enables SIGSEGV reception to the SKAS sig_handler_common, which matches the tt code. I still need to figure out why the SA_NODEFER flag was backed out in favor of this. Signed-off-by:
-
Paolo \'Blaisorblade\' Giarrusso authored
Change some config text (hide CONFIG_MODVERSION which is broken on UML and fix a dummy prompt). Signed-off-by:
-
Paolo \'Blaisorblade\' Giarrusso authored
Re-add some needed headers inclusion deleted in http://linux.bkbits.net:8080/linux-2.5/cset@41e49628dGbOWX-bT9yZII4f19GT6A If you think it cannot make sense to include both <sys/ptrace.h> and <linux/ptrace.h> (as userspace process, i.e. host includes), go complaining with glibc, or follow the linux-abi includes idea. However, the compilation failure is possibly glibc-version (or better glibc includes version) related - what I now is that the failure happens on my system with a glibc 2.3.4 (from Gentoo). Also, fix the syscall table to both compile and have no empty slot (which could cause Oopses). Acked-by:
Jeff Dike <jdike@addtoit.com> Signed-off-by:
-
Paolo \'Blaisorblade\' Giarrusso authored
arch/um/Kconfig_arch is actually a symlink, so * Remove it from the tree. * Make sure it is removed during make mrproper. Signed-off-by:
-
Andrew Morton authored
With NFSD=y, EXPORTFS=m: fs/built-in.o(.text+0x7fff5): In function `fh_verify': /usr/src/25/fs/nfsd/nfsfh.c:221: undefined reference to `export_op_default' So it seems that nfsd requires exportfs. For consistency we should tidy up the definition of EXPORTFS too. Signed-off-by:
Neil Brown <neilb@cse.unsw.edu.au> Signed-off-by:
-
Michal Ludvig authored
Signed-off-by:
Michal Ludvig <michal@logix.cz> Signed-off-by:
-
Nishanth Aravamudan authored
Fixes the math of both jiffies_to_usecs() and usecs_to_jiffies() which improperly assume the same rounding point -- 1,000 -- as jiffies_to_msecs() and msecs_to_jiffies(), when in fact it should be 1,000,000. Furthermore, the actual math of both functions is actually wrong and will lead to more than just rounding errors. Signed-off-by:
Nishanth Aravamudan <nacc@us.ibm.com> Signed-off-by:
-
Jimi Xenidis authored
The hvcs driver does not register a devfs_name resulting in devfs creating /dev/<NULL>* entries. The following one line patch remedies the problem. Signed-off-by:
Jimi Xenidis <jimix@watson.ibm.com> Signed-off-by:
-
Dave Jones authored
Yet another ppc64 build failure.. Move the function before its first usage, and the failure goes away. Signed-off-by:
Dave Jones <davej@redhat.com> Signed-off-by:
-
Zou Nanhai authored
There is an obvious error in the header of /proc/slabinfo Signed-off-by:
Zou Nan hai <nanhai.zou@intel.com> Signed-off-by:
-
Josh Green authored
A wrong ordering of operators in pcmica_register_client() causes initialization problems with multiple PCMCIA cards. Signed-off-by:
Josh Green <jgreen@users.sourceforge.net> Signed-off-by:
Dominik Brodowski <linux@dominikbrodowski.de> Signed-off-by:
-
Stephen D. Smalley authored
This fixes the selinux_inode_setattr hook function to honor the ATTR_FORCE flag, skipping any permission checking in that case. Otherwise, it is possible though unlikely for a denial from the hook to prevent proper updating, e.g. for remove_suid upon writing to a file. This would only occur if the process had write permission to a suid file but lacked setattr permission to it. Signed-off-by:
Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by:
James Morris <jmorris@redhat.com> Signed-off-by:
-
Greg Kroah-Hartman authored
I have a new job. Signed-off-by:
Greg Kroah-Hartman <greg@kroah.com> Signed-off-by:
-
David Howells authored
This cleans up a comment in vmlinux.lds.S - emacs now has an LD script mode, so it shouldn't be forced into C mode. Signed-Off-By:
David Howells <dhowells@redhat.com> Signed-off-by:
-
David Howells authored
This makes the bit finding functions in asm/bitops.h take const pointers since they don't modify what they access. Signed-Off-By:
-
David Howells authored
This adds TIF_MEMDIE for FRV. Could whoever added it to include/asm-i386/thread_info.h comment this flag there please? I've given it a comment here, but I'm not sure it's correct. Signed-Off-By:
-
David Howells authored
This fixes a race in the FRV arch's semaphore implementation. The same type of fixes were applied to the rw-semaphore implementations to fix the same races there. The race involved the on-stack record linked into the semaphore's queue by the down() executed by a process now sleeping on the semaphore going away and the sleeping task going away before the process that woke it up during up() processing had finished with those structures. Signed-Off-By:
-
David Howells authored
This cleans up the remaining references to the cli() and sti() functions from the FRV arch now they're deprecated. Signed-Off-By:
-
David Howells authored
This makes switch_to() on the FRV pass through and return the previous task pointer rather than trusting to luck that it'll be left in the correct register/variable. Signed-Off-By:
-
