The seccomp check has to happen when entering the syscall and not when
exiting it or regs->gpr[0] contains garabge during signal handling in
ppc64_rt_sigreturn (this actually might be a bug too, but an orthogonal
one, since we really have to run the check before invoking the syscall and
not after it).
Signed-off-by: Andrea Arcangeli <andrea@cpushare.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch includes ppc64 architecture specific changes to support temporary
disarming on reentrancy of probes.
Signed-of-by: Prasanna S Panchamukhi <prasanna@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The architecture independent code of the current kprobes implementation is
arming and disarming kprobes at registration time. The problem is that the
code is assuming that arming and disarming is a just done by a simple write
of some magic value to an address. This is problematic for ia64 where our
instructions look more like structures, and we can not insert break points
by just doing something like:
*p->addr = BREAKPOINT_INSTRUCTION;
The following patch to 2.6.12-rc4-mm2 adds two new architecture dependent
functions:
* void arch_arm_kprobe(struct kprobe *p)
* void arch_disarm_kprobe(struct kprobe *p)
and then adds the new functions for each of the architectures that already
implement kprobes (spar64/ppc64/i386/x86_64).
I thought arch_[dis]arm_kprobe was the most descriptive of what was really
happening, but each of the architectures already had a disarm_kprobe()
function that was really a "disarm and do some other clean-up items as
needed when you stumble across a recursive kprobe." So... I took the
liberty of changing the code that was calling disarm_kprobe() to call
arch_disarm_kprobe(), and then do the cleanup in the block of code dealing
with the recursive kprobe case.
So far this patch as been tested on i386, x86_64, and ppc64, but still
needs to be tested in sparc64.
Signed-off-by: Rusty Lynch <rusty.lynch@intel.com>
Signed-off-by: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The attached patch causes the various arch specific install.sh scripts to
look for ${CROSS_COMPILE}installkernel rather than just installkernel (in
both /sbin/ and ~/bin/ where the script already did this). This allows you
to have e.g. arm-linux-installkernel as a handy way to install on your
cross target. It also prevents the script picking up on the host
/sbin/installkernel which causes the script to fall through and do the
install itself (which is what I actually use myself, with $INSTALL_PATH
set).
I don't believe it causes back-compatibility problems since calling the
host installkernel was never likely to work or be what you wanted when
cross compiling anyway. If $CROSS_COMPILE isn't set then nothing changes.
I only use ARM and i386 myself but I figured it couldn't hurt to do the
whole lot. I've cc'd those who I hope are the arch maintainers for files
that I've touched.
Signed-off-by: Ian Campbell <icampbell@arcom.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Provide the architecture specific implementation for SPARSEMEM for PPC64
systems.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Mike Kravetz <kravetz@us.ibm.com> (in part)
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Provide hooks for PPC64 to allow memory models to be informed of installed
memory areas. This allows SPARSEMEM to instantiate mem_map for the populated
areas.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Provide an implementation of early_pfn_to_nid for PPC64. This is used by
memory models to determine the node from which to take allocations before the
memory allocators are fully initialised.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The part of the sparsemem patch which modifies memmap_init_zone() has recently
become a problem. It changes behavior so that there is a call to
pfn_to_page() for each individual page inside of a node's range:
node_start_pfn through node_end_pfn. It used to simply do this once, at the
beginning of the node, but having sparsemem's non-contiguous mem_map[]s inside
of a node made it necessary to change.
Mike Kravetz recently wrote a patch which made the NUMA code accept some new
kinds of layouts. The system's memory was laid out like this, with node 0's
memory in two pieces: one before and one after node 1's memory:
Node 0: +++++ +++++
Node 1: +++++
Previous behavior before Mike's patch was to assign nodes like this:
Node 0: 00000 XXXXX
Node 1: 11111
Where the 'X' areas were simply thrown away. The new behavior was to make the
pg_data_t span node 0 across all of its areas, including areas that are really
node 1's: Node 0: 000000000000000 Node 1: 11111
This wastes a little bit of mem_map space, but ends up being OK, and more
fully utilizes the system's memory. memmap_init_zone() initializes all of the
"struct page"s for node 0, even for the "hole", but those never get used,
because there is no pfn_to_page() that resolves to those pages. However, only
calling pfn_to_page() once, memmap_init_zone() always uses the pages that were
allocated for node0->node_mem_map because:
struct page *start = pfn_to_page(start_pfn);
// effectively start = &node->node_mem_map[0]
for (page = start; page < (start + size); page++) {
init_page_here();...
page++;
}
Slow, and wasteful, but generally harmless.
But, modify that to call pfn_to_page() for each loop iteration (like sparsemem
does):
for (pfn = start_pfn; pfn < < (start_pfn + size); pfn++++) {
page = pfn_to_page(pfn);
}
And you end up trying to initialize node 1's pages too early, along with bogus
data from node 0. This patch checks for those weird layouts and declines to
touch the pages, making the more frequent pfn_to_page() calls OK to do.
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch changes some of the default behavior in the ppc64 Kconfig file
that was recently changed/added to 2.6.12-rc2-mm1 by Dave Hansen in
preparation for SPARSEMEM. Patch allows the display of both FLAT and
DISCONTIG models on pseries. As before, default is DISCONTIG for SMP and
PSERIES and FLAT for others.
Signed-off-by: Mike Kravetz <kravetz@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This will at least suppress one prompt that users would have received the
first time they compile with the new DISCONTIG arch option. They'll still
get the "Memory Model" prompt, but 99% of them will have the default work
there.
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
For all architectures, this just means that you'll see a "Memory Model"
choice in your architecture menu. For those that implement DISCONTIGMEM,
you may eventually want to make your ARCH_DISCONTIGMEM_ENABLE a "def_bool
y" and make your users select DISCONTIGMEM right out of the new choice
menu. The only disadvantage might be if you have some specific things that
you need in your help option to explain something about DISCONTIGMEM.
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch effectively eliminates direct use of pgdat->node_mem_map outside
of the DISCONTIG code. On a flat memory system, these fields aren't
currently used, neither are they on a sparsemem system.
There was also a node_mem_map(nid) macro on many architectures. Its use
along with the use of ->node_mem_map itself was not consistent. It has
been removed in favor of two new, more explicit, arch-independent macros:
pgdat_page_nr(pgdat, pagenr)
nid_page_nr(nid, pagenr)
I called them "pgdat" and "nid" because we overload the term "node" to mean
"NUMA node", "DISCONTIG node" or "pg_data_t" in very confusing ways. I
believe the newer names are much clearer.
These macros can be overridden in the sparsemem case with a theoretically
slower operation using node_start_pfn and pfn_to_page(), instead. We could
make this the only behavior if people want, but I don't want to change too
much at once. One thing at a time.
This patch removes more code than it adds.
Compile tested on alpha, alpha discontig, arm, arm-discontig, i386, i386
generic, NUMAQ, Summit, ppc64, ppc64 discontig, and x86_64. Full list
here: http://sr71.net/patches/2.6.12/2.6.12-rc1-mhp2/configs/
Boot tested on NUMAQ, x86 SMP and ppc64 power4/5 LPARs.
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin J. Bligh <mbligh@aracnet.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Currently reset and powerdown are not implemented on the Maple board,
and attempting to do so will (incorrectly return). This implements
the proper communication with the service processor, allowing correct
reset and powerdown on the Maple board, by communicating with the
service processor. If somehow it's unable to communicate with the
service processor it will loop forever instead.
Note that powerdown on the Maple will power down the CPUs, but not the
fans or other board components due to hardware and firmware
limitations.
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Frank Rowand <frowand@mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
For I/O DLPAR to work properly, the kernel needs to allow for dynamic
assignment of the irq field of the pci_dev structure upon dynamic bus
addition. This patch moves the assignment of that field from
pSeries_final_fixup() to pcibios_fixup_bus(), which enables dynamic
assignment for the children of a newly added bus.
Currently, pci_devs receive their irq numbers in one of two ways. The
irq line is either read at boot for all pci_devs, or read by the rpaphp
module at slot enable time. The latter is no longer sufficient for
DLPAR addition of slots that don't qualify as PCI-hotplug capable.
This solution handles the cases of boot and dynamic add.
Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
This patch corrects the printing of progress indicators to the op
panel on p/iSeries ppc64 systems. Each discrete reference code should
begin with a form feed char to clear the op panel, and the first and
second lines should be separated with a CR/LF sequence. Padding with
spaces is not necessary.
Also, capitalize the hex value printed on the first line, to be
consistent with the values printed by firmware, service processor,
etc.
It turns out that there's an ibm,form-feed property; this patch uses
it in the pSeries-specific progress routine. This patch also checks
the number of rows and the specific width of each row (the second row
on power5 systems can actually hold 80 characters). If the displayed
text is too wide for the physical display, it can be viewed in the ASM
menus, or by selecting option 14 on the op panel.
Signed-off-by: Mike Strosaker <strosake@austin.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Implementation of software load support for the BE iommu. This is very
different from other iommu code on ppc64, since we only do a static mapping.
The mapping is currently hardcoded but should really be read from the
firmware, but they don't set up the device nodes yet. There is a single
512MB DMA window for PCI, USB and ethernet at 0x20000000 for our RAM.
The Cell processor can put the I/O page table either in memory like
the hashed page table (hardware load) or have the operating system
write the entries into memory mapped CPU registers (software load).
I use the software load mechanism because I know that all I/O page
table entries for the amount of installed physical memory fit into
the IO TLB cache. At the point when we get machines with more than
4GB of installed memory, we can either use hardware I/O page table
access like the other platforms do or dynamically update the I/O
TLB entries when a page fault occurs in the I/O subsystem.
The software load can then use the macros that I have implemented
for the static mapping in order to do the TLB cache updates.
Signed-off-by: Arnd Bergmann <arndb@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Add support for the integrated interrupt controller on BPA
CPUs. There is one of those for each SMT thread.
The mapping of interrupt numbers to HW interrupt sources
is described in arch/ppc64/kernel/bpa_iic.h.
This version hardcodes the 'Spider' chip as the secondary
interrupt controller. That is not really generic for the
architecture, but at the moment it is the only secondary
PIC that exists.
A little more work will be needed on this as soon as
we have boards with multiple external interrupt controllers.
Signed-off-by: Arnd Bergmann <arndb@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
This adds the basic support for running on BPA machines.
So far, this is only the IBM workstation, and it will
not run on others without a little more generalization.
It should be possible to configure a kernel for any
combination of CONFIG_PPC_BPA with any of the other
multiplatform targets.
Signed-off-by: Arnd Bergmann <arndb@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
The firmware provides the location and size of the nvram
in the device tree, so it does not really contain any
hardware specific bits and could be used on other
machines as well.
Signed-off-by: Arnd Bergmann <arndb@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
The pSeries_progress function is called from some places in the rtas code,
which may also be used by non-pSeries platforms.
Though pSeries is currently the only platform type that implements
display-character, the code is actually generic enough to be part of
the rtas subsystem.
I hit a bug here because the generic rtas code tried calling ppc_md.progress,
which points to an __init function on most platforms.
We could also clear the ppc_md.progress pointer when freeing the init memory
to make it more explicit that ppc_md.progress must not be called after
bootup.
Signed-off-by: Arnd Bergmann <arndb@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
BPA is using rtas for PCI but should not be confused by
pSeries code. This also avoids some #ifdefs. Other
platforms that want to use rtas_pci.c could create
their own platform_pci.c with platform specific fixups.
Signed-off-by: Arnd Bergmann <arndb@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
The rtc rtas functions are not pSeries specific but can
also be used by BPA and other SLOF based platforms
Signed-off-by: Arnd Bergmann <arndb@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
pSeries and maple have almost the same code for calibrate_decr,
and BPA would need yet another copy. Instead, I'm moving the
code to arch/ppc64/kernel/time.c.
Some of the related declarations were missing from header
files, so I'm moving those as well.
It makes sense to merge this with the pmac function of the
same name, so we end up having just one implemetation for
iSeries and one for Open Firmware based machines.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Allow the SMT bit to be set/reset at boot, like the ALTIVEC bit. This
means we will enable SMT on unknown cpus that support it.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We dont use the hardware referenced and changed bits and setting them early
avoids a store to memory. We already do this for userspace hptes but not
kernel ones. Do it.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Currently we dynamically allocate the fake parent device for all devices on
the vio bus. This patch statically allocates it. This also allows us to
reuse it for the iSeries "generic" vio device (that is used for passing to
dma routines when communicating with the hypervisor without a device
involved). Also unexport vio_bus_type as it is never used in modules.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch allows iSeries to build with CONFIG_PCI=n. This is useful for
partitions that have only virtual I/O.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch just removes some dead code, fixes messages that referred to the
file this code used to be in and inserts XmPciLpEvent_init into its caller.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch just merges XmPciLpEvent.c into iSeries_irq.c (the only caller of
its only external function). XmPciLpEvent.c just contained the lowlevel
iSeries irq code.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch is just simple cleanups to the iSeries irq code.
- whitespace and comments
- rearrange some functions to avoid forward declarations
- remove XmPciLpEvent.h as its functions were declared elsewhere
- remove decaration of function that no longer exists
No semantic changes.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The AgentId, PhbId, FrameId, CardLocation and Location members of
iSeries_Device_Node are stored early in the boot process just so that a
message about the device can be printed later in the boot process. Remove
them and construct the message by doing the VPD parsing at the time the
message is printed.
Also remove a few unused defines in iSeries_VpdInfo.c.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The IoRetry member of iSeries_Devide_Node is really only used locally, so
remove it and replace it with a local variable.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Remove no longer used things from iSeries_pci.h.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Clean up iSeries_VpdInfo.c:
- white space and comment fixes
- make a function static
- the functions here are only called from iSeries_pci.c, so
CONFIG_PCI will be set (so remove check)
- only build when CONFIG_PCI is set
- remove unneeded includes and cast
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The file arch/ppc64/kernel/iSeries_pci_reset contains only one function that
is not use anywhere (any more). Remove it. This function is the only user of
the ReturnCode member of iSeries_Device_Node, so remove that as well.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Last of this round of the iSeries header cleanups
- don't have two defines for the same thing (HvMaxArchitectedLps
and HvMaxArchitectedVirtualLans)
- HvCallSc.h only needs linux/types.h
- remove unused struct definition
- add "extern" to some more function declarations
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch removes some unused bits from HvCall.h and some unneeded #includes
from other files. Also includes ItLpQueue.h in paca.h in preference to a stub
declaration of struct ItLpQueue.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Just white space cleaups and move process_iSeries_events into its only caller.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch removes from the iSeries header files a large number of inline
functions that are not used. It also changes the only caller of a HvCallCfg
function that is outside HvLpConfig.h to its equivalent HvLpConfig function
and no longer includes HvCallCfg.h where it is not needed.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
include/asm-ppc64/iSeries/LparData.h just included a whole lot of other files
to declare variables that would be better declared in those other files. So,
remove it. This will reduce that number of things needed to be included in
most cases to access the relevant variables.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
include/asm-ppc64/iSeries/iSeries_proc.h just contains a declaration of a
function that no longer exists. Remove it.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The following kind of calls currently fails :
make ARCH=ppc64 CC="gcc-3.4"
Since the code for detecting a biarch compiler and adding the needed 64bit
magic argument fails if the AS/LD/CC commands are overriden in the command
line.
The attached patch fixes this by using the make override and += directive,
but i am not 100% sure this will work without gmake, as i am no Makefile
expert.
Cc: Paul Mackerras <paulus@samba.org>
Cc: Anton Blanchard <anton@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Currently ppc64 has two mm_structs for the kernel, init_mm and also
ioremap_mm. The latter really isn't necessary: this patch abolishes it,
instead restricting vmallocs to the lower 1TB of the init_mm's range and
placing io mappings in the upper 1TB. This simplifies the code in a number
of places and eliminates an unecessary set of pagetables. It also tweaks
the unmap/free path a little, allowing us to remove the unmap_im_area() set
of page table walkers, replacing them with unmap_vm_area().
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch kills the whole embedded System.map mecanism and the
bootloader-passed System.map that was used to provide symbol resolution in
xmon. Instead, xmon now uses kallsyms like ppc64 does.
No hurry getting that in Linus tree, let it be tested in -mm for a while
first and make sure it doesn't break various embedded configs.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Ingo recently introduced a great speedup for allocating new mmaps using the
free_area_cache pointer which boosts the specweb SSL benchmark by 4-5% and
causes huge performance increases in thread creation.
The downside of this patch is that it does lead to fragmentation in the
mmap-ed areas (visible via /proc/self/maps), such that some applications
that work fine under 2.4 kernels quickly run out of memory on any 2.6
kernel.
The problem is twofold:
1) the free_area_cache is used to continue a search for memory where
the last search ended. Before the change new areas were always
searched from the base address on.
So now new small areas are cluttering holes of all sizes
throughout the whole mmap-able region whereas before small holes
tended to close holes near the base leaving holes far from the base
large and available for larger requests.
2) the free_area_cache also is set to the location of the last
munmap-ed area so in scenarios where we allocate e.g. five regions of
1K each, then free regions 4 2 3 in this order the next request for 1K
will be placed in the position of the old region 3, whereas before we
appended it to the still active region 1, placing it at the location
of the old region 2. Before we had 1 free region of 2K, now we only
get two free regions of 1K -> fragmentation.
The patch addresses thes issues by introducing yet another cache descriptor
cached_hole_size that contains the largest known hole size below the
current free_area_cache. If a new request comes in the size is compared
against the cached_hole_size and if the request can be filled with a hole
below free_area_cache the search is started from the base instead.
The results look promising: Whereas 2.6.12-rc4 fragments quickly and my
(earlier posted) leakme.c test program terminates after 50000+ iterations
with 96 distinct and fragmented maps in /proc/self/maps it performs nicely
(as expected) with thread creation, Ingo's test_str02 with 20000 threads
requires 0.7s system time.
Taking out Ingo's patch (un-patch available per request) by basically
deleting all mentions of free_area_cache from the kernel and starting the
search for new memory always at the respective bases we observe: leakme
terminates successfully with 11 distinctive hardly fragmented areas in
/proc/self/maps but thread creating is gringdingly slow: 30+s(!) system
time for Ingo's test_str02 with 20000 threads.
Now - drumroll ;-) the appended patch works fine with leakme: it ends with
only 7 distinct areas in /proc/self/maps and also thread creation seems
sufficiently fast with 0.71s for 20000 threads.
Signed-off-by: Wolfgang Wander <wwc@rentec.com>
Credit-to: "Richard Purdie" <rpurdie@rpsys.net>
Signed-off-by: Ken Chen <kenneth.w.chen@intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu> (partly)
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
A lot of the code in arch/*/mm/hugetlbpage.c is quite similar. This patch
attempts to consolidate a lot of the code across the arch's, putting the
combined version in mm/hugetlb.c. There are a couple of uglyish hacks in
order to covert all the hugepage archs, but the result is a very large
reduction in the total amount of code. It also means things like hugepage
lazy allocation could be implemented in one place, instead of six.
Tested, at least a little, on ppc64, i386 and x86_64.
Notes:
- this patch changes the meaning of set_huge_pte() to be more
analagous to set_pte()
- does SH4 need s special huge_ptep_get_and_clear()??
Acked-by: William Lee Irwin <wli@holomorphy.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch implements a number of smp_processor_id() cleanup ideas that
Arjan van de Ven and I came up with.
The previous __smp_processor_id/_smp_processor_id/smp_processor_id API
spaghetti was hard to follow both on the implementational and on the
usage side.
Some of the complexity arose from picking wrong names, some of the
complexity comes from the fact that not all architectures defined
__smp_processor_id.
In the new code, there are two externally visible symbols:
- smp_processor_id(): debug variant.
- raw_smp_processor_id(): nondebug variant. Replaces all existing
uses of _smp_processor_id() and __smp_processor_id(). Defined
by every SMP architecture in include/asm-*/smp.h.
There is one new internal symbol, dependent on DEBUG_PREEMPT:
- debug_smp_processor_id(): internal debug variant, mapped to
smp_processor_id().
Also, i moved debug_smp_processor_id() from lib/kernel_lock.c into a new
lib/smp_processor_id.c file. All related comments got updated and/or
clarified.
I have build/boot tested the following 8 .config combinations on x86:
{SMP,UP} x {PREEMPT,!PREEMPT} x {DEBUG_PREEMPT,!DEBUG_PREEMPT}
I have also build/boot tested x64 on UP/PREEMPT/DEBUG_PREEMPT. (Other
architectures are untested, but should work just fine.)
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Andrea Arcangeli