- 27 Jun, 2004 22 commits
-
-
Rusty Russell authored
store_stackinfo() does an unlocked module list walk during normal runtime which opens up a race with the module load/unload code. This can be triggered by simply unloading and loading a module in a loop with CONFIG_DEBUG_PAGEALLOC resulting in store_stackinfo() tripping over bad list pointers. kernel_text_address doesn't take any locks, because during an OOPS we don't want to deadlock. Rename that to __kernel_text_address, and make kernel_text_address take the lock. Signed-off-by:
Zwane Mwaikambo <zwane@fsmlabs.com> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (modified) Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Martin J. Bligh authored
I find __free_pages_bulk very hard to understand ... (I was trying to mod it for the non MAX_ORDER aligned zones, and cleaned it up first). This should make it much more comprehensible to mortal man ... I benchmarked the changes on the big 16x and it's no slower (actually it's about 0.5% faster, but that's within experimental error). I moved the creation of mask into __free_pages_bulk from the caller - it seems to really belong inside there. Then instead of doing wierd limbo dances with mask, I made it use order instead where it's more intuitive. Personally I find this makes the whole thing a damned sight easier to understand. Signed-off-by:
-
Coywolf Qi Hunt authored
I just find a bug that ``make distclean'' cannot remove the editor backup files and the like when using build directory. That is because the find command is improperly searching the build directory instead of the $(srctree) it should. Signed-off-by:
-
Keith Owens authored
Verify that linking kallsyms into vmlinux generates a stable System.map, instead of assuming that it is stable. Add CONFIG_KALLSYMS_EXTRA_PASS as a temporary workaround for unstable maps, so users can proceed while waiting for kallsyms to be fixed. Signed-off-by:
Keith Owens <kaos@sgi.com> Signed-off-by:
-
Keith Owens authored
Exclude symbols added by kallsyms itself, so .tmp_kallsyms[12].S have the same list of symbols. Signed-off-by:
-
Zwane Mwaikambo authored
This is a small cleanup requested by Urban, use the rcls/err in smb_request as opposed to smb_sb_info. Signed-off-by:
Zwane Mwaikambo <zwane@linuxpower.ca> Signed-off-by:
-
Andrew Morton authored
This BUG_ON() triggers for `count = -EFOO' due to PAGE_SIZE being unsigned. Signed-off-by:
-
Ingo Molnar authored
we'd like to announce the availability of the following kernel patch: http://redhat.com/~mingo/nx-patches/nx-2.6.7-rc2-bk2-AE which makes use of the 'NX' x86 feature pioneered in AMD64 CPUs and for which support has also been announced by Intel. (other x86 CPU vendors, Transmeta and VIA announced support as well. Windows support for NX has also been announced by Microsoft, for their next service pack.) The NX feature is also being marketed as 'Enhanced Virus Protection'. This patch makes sure Linux has full support for this hardware feature on x86 too. What does this patch do? The pagetable format of current x86 CPUs does not have an 'execute' bit. This means that even if an application maps a memory area without PROT_EXEC, the CPU will still allow code to be executed in this memory. This property is often abused by exploits when they manage to inject hostile code into this memory, for example via a buffer overflow. The NX feature changes this and adds a 'dont execute' bit to the PAE pagetable format. But since the flag defaults to zero (for compatibility reasons), all pages are executable by default and the kernel has to be taught to make use of this bit. If the NX feature is supported by the CPU then the patched kernel turns on NX and it will enforce userspace executability constraints such as a no-exec stack and no-exec mmap and data areas. This means less chance for stack overflows and buffer-overflows to cause exploits. furthermore, the patch also implements 'NX protection' for kernelspace code: only the kernel code and modules are executable - so even kernel-space overflows are harder (in some cases, impossible) to exploit. Here is how kernel code that tries to execute off the stack is stopped: kernel tried to access NX-protected page - exploit attempt? (uid: 500) Unable to handle kernel paging request at virtual address f78d0f40 printing eip: ... The patch is based on a prototype NX patch written for 2.4 by Intel - special thanks go to Suresh Siddha and Jun Nakajima @ Intel. The existing NX support in the 64-bit x86_64 kernels has been written by Andi Kleen and this patch is modeled after his code. Arjan van de Ven has also provided lots of feedback and he has integrated the patch into the Fedora Core 2 kernel. Test rpms are available for download at: http://redhat.com/~arjanv/2.6/RPMS.kernel/ the kernel-2.6.6-1.411 rpms have the NX patch applied. here's a quickstart to recompile the vanilla kernel from source with the NX patch: http://redhat.com/~mingo/nx-patches/QuickStart-NX.txt update: - make the heap non-executable on PT_GNU_STACK binaries. - make all data mmap()s (and the heap) executable on !PT_GNU_STACK (legacy) binaries. This has no effect on non-NX CPUs, but should be much more compatible on NX CPUs. The only effect it has it has on non-NX CPUs is the extra 'x' bit displayed in /proc/PID/maps. Signed-off-by: -
Stas Sergeev authored
The previous discussion was started here: http://www.uwsg.iu.edu/hypermail/linux/kernel/0211.0/0477.html but in 2.4 times this was kind of problematic. Now, with the lazy bitmap allocation and per-CPU TSS, this will really not drain any resources I think. 8K TSS increase and 8K per process *that does ioperm()* - I think it is not very bad. The reasons why I need that, are described in the URL above. Basically this will allow to use full-screen VESA under dosemu (without LFB though), and this may be also helpfull for the XFree project and some other projects: http://www.uwsg.iu.edu/hypermail/linux/kernel/9807.1/1079.htmlSigned-off-by:
-
Anton Blanchard authored
__alloc_bootmem_node currently panics if it cant satisfy an allocation for a particular node. Thats rather antisocial, we should at the very least return NULL and allow the caller to proceed (eg try another node). A quick look at alloc_bootmem_node usage suggests we should fall back to allocating from other nodes if it fails (as arch/alpha/kernel/pci_iommu.c and arch/x86_64/kernel/setup64.c do). The following patch does that. We fall back to the regular __alloc_bootmem when __alloc_bootmem_node fails, which means all other nodes are checked for available memory. Signed-off-by:
Anton Blanchard <anton@samba.org> Signed-off-by:
-
David Howells authored
This fixes a bit of iSeries code that hadn't been changed to reflect that cpumask_t is now a struct and not an integer type. Signed-off-by:
-
Anton Blanchard authored
rtas.c doesn't call reloc_offset remove comment from rtas.c and extern function declaration Signed-off-by:
Milton Miller <miltonm@bga.com> Signed-off-by:
-
Anton Blanchard authored
Some udbg.c cleanups: - remove some old comments - clean up formatting - remove unused udbg_puthex and udbg_printSP Signed-off-by:
-
Anton Blanchard authored
udbg_read would lock up if ppc_md.udbg_putc didnt exist. Just return 0 and dont lock up instead. Signed-off-by:
-
Anton Blanchard authored
Use vsnprintf/snprintf in udbg.c Signed-off-by:
-
Anton Blanchard authored
- Use 970/970FX instead of GPUL - Add POWER5 and 970FX to systemcfg.h - Create new cpu feature CPU_FTR_MMCRA_SIHV and use it Signed-off-by:
-
Andrew Morton authored
In file included from arch/ppc64/kernel/setup.c:37: include/asm/machdep.h:116: `COMMAND_LINE_SIZE' undeclared here (not in a function) Signed-off-by:
-
Bjorn Helgaas authored
Add support for the EFI/DIG PCDP console discovery table (see http://www.dig64.org/specifications/DIG64_HCDPv20_042804.pdf). This moves the code from drivers/serial/8250_hcdp.[ch] to drivers/firmware/pcdp.[ch], since it's no longer 8250-specific. It also obsoletes CONFIG_SERIAL_8250_HCDP, replacing it with CONFIG_EFI_PCDP (which defaults to Y for ia64). In a nutshell, HCDP tells us "these UARTs are available for use as a console," and it's up to the user to explicitly specify the console device. The kernel can guess in some cases, but not all. The PCDP (aka HCDP v2) tells us what we really want to know, namely, "this UART or VGA device is the console device." (It also has provision for support for new device types.) Signed-off-by:
Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by:
-
bk://linux-scsi.bkbits.net/scsi-for-linus-2.6Linus Torvalds authored
into ppc970.osdl.org:/home/torvalds/v2.6/linux
-
Alan Stern authored
On Wed, 23 Jun 2004, Mike Anderson wrote: > Since SCSI already has the device list it would seem like we would > possibly add a new flag like was done for mode sense. As this is not a > transport issue I would assume we would not want to add flags in > usb/storage, but handle it in the mid-layer as a SCSI protocol > non-compliance. Here is a patch that implements this suggestion. It's rather similar to the one that Javier Marcet wrote back in January, in http://marc.theaimsgroup.com/?l=linux-usb-users&m=107345268526718&w=2Signed-off-by:
Alan Stern <stern@rowland.harvard.edu> Signed-off-by:
James Bottomley <James.Bottomley@SteelEye.com>
-
Guennadi Liakhovetski authored
Signed-off-by: -
Christoph Hellwig authored
On Sun, Jun 06, 2004 at 02:41:56PM +0200, Christoph Hellwig wrote: > I've looked through my old tmscsim patch queue and found this one: > > - merge dc390_initDCB into dc390_slave_alloc > - merge DC390_release and dc390_shutdown into dc390_remove_one, > use del_timer_sync to make sure the timer is really deleted on > removal, adjust locking accordingly > - some tiny related cleanups Okay, here's a resend vs current scsi-misc-2.6 that has your three outstanding merged. Additionally I've also killed dc390_freeDCBs() as all dcbs are removed in ->slave_destroy. Signed-off-by:
-
- 26 Jun, 2004 18 commits
-
-
James Bottomley authored
Since the driver now supports multiple virtual segments in an iommu situation, it can also support clustering for non-iommu systems. Signed-off-by: -
Andrew Morton authored
drivers/message/fusion/mptbase.c: In function `procmpt_summary_read': drivers/message/fusion/mptbase.c:5160: parse error before `int' Signed-off-by:
-
James Bottomley authored
Signed-off-by: -
James Bottomley authored
-
James Bottomley authored
aic7xxx probing routines are still wrong on eisa. Fix eisa by incrementing found if it returns successfully Also make all the various incarnations of the pci probing routine consistently return the number of found cards (or 1 for the later generic device model probing). Signed-off-by: -
James Bottomley authored
-
Randy Dunlap authored
| From: Andrew Morton | To: linux-scsi@vger.kernel.org | Subject: fdomain screwup | | - fdomain_setup() is marked __init, but is called from non-__init | fdomain_16x0_detect() | | - fdomain_setup() is declared in drivers/scsi/pcmcia/fdomain_stub.c as | taking two arguments, but is implemented in fdomain_setup() taking a | single argument. | | Please, never ever ever put extern function declarations in .c files. | Put it in a header file which is visible to the definition and to all | callsites. | | - fdomain_setup() is declared static, hence the linkage fails. | | - fdomain_16x0_bus_reset() is implemented in drivers/scsi/fdomain.c but | has static scope, so the call from drivers/scsi/pcmcia/fdomain_stub.c | doesn't work. [I didn't find this to be the case, so I didn't fix it.] | - fdomain_16x0_bus_reset() has an extern declaration in | drivers/scsi/pcmcia/fdomain_stub.c. This should be moved to a header | file which is visible to etc... Builds and loads. No hardware to test. Signed-off-by:
Randy Dunlap <rddunlap@osdl.org> Signed-off-by:
-
bk://bk.arm.linux.org.uk/linux-2.6-rmkLinus Torvalds authored
into ppc970.osdl.org:/home/torvalds/v2.6/linux
-
Russell King authored
-
bk://linux-mtd.bkbits.net/sbc85xx-2.6Linus Torvalds authored
into ppc970.osdl.org:/home/torvalds/v2.6/linux
-
Linus Torvalds authored
-
Christoph Hellwig authored
On Mon, Jun 21, 2004 at 12:16:08PM -0400, Moore, Eric Dean wrote: > We are pleased to announce the MPT Fusion release candidate for lk 2.6 I've worked with Eric offline to resolve the issue we had and get some more changes in, but he has left for his vacation today. He send me a patch though and left it to me whether we'd merged it despite only moderate testing. Given 2.6.7 was just done and he'll certainly be back before 2.6.8 I'd go for it. Below is the patch rediffed against scsi-misc-2.6: Signed-off-by: -
bk://ppc@ppc.bkbits.net/for-linus-ppcTom Rini authored
into kernel.crashing.org:/home/trini/work/kernel/pristine/for-linus-ppc
-
James Bottomley authored
The DMA conversion of the advansys driver is still broken. Add a #warning to the driver and a comment above it explaining what needs to be done. Mark the driver as BROKEN because of the warning Also remove the #include "scsi.h" Signed-off-by: -
Douglas Gilbert authored
- fix PCI wide "use_sg > 0" problem introduced in lk 2.6.7-bk2 - fix ISA oops (present in all versions of the lk 2.6 series to date) - bump version to "3.3K" and add comments - fix compilation error when ADVANSYS_DEBUG defined Signed-off-by: -
Christoph Hellwig authored
PPC has an out of line and exported abs() that gives lots of nice and wierd compilation erorrs. Also kill the duplicate cpu_online() in asm-ppc/smp.h.
-
bk://bk.arm.linux.org.uk/linux-2.6-pcmciaLinus Torvalds authored
into ppc970.osdl.org:/home/torvalds/v2.6/linux
-
Russell King authored
0 is not a valid alignment value for allocate_resource()
-
