Matt Evans [Tue, 31 Jan 2012 06:30:41 +0000 (17:30 +1100)]
kvm tools: Repair running on non ioeventfd-capable platforms
Commit d3923126a24212f1e746a84a575dadbd9f259418 added a bunch of
nice error checking around ioevent__init() but the init may gracefully
fail if ioevents simply aren't supported (PPC64 KVM).
This commit adds a new return code for the init -- positive, but
identifiable as 'not success 0'.
Signed-off-by: Matt Evans <matt@ozlabs.org> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Commit b50cac5 ("PCI/sysfs: add per pci device msi[x] irq listing (v5)") added
"struct kobject" dependency to <linux/msi.h> which breaks KVM tool build. As
the header file is no longer about MSI hardware, make a new header file for
"struct msi_ms".
Reported-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Ingo Molnar [Wed, 18 Jan 2012 09:04:48 +0000 (10:04 +0100)]
kvm tools: Fix build breakage with GCC 4.7
* Pekka Enberg <penberg@kernel.org> wrote:
> >In file included from virtio/net.c:3:0:
> >include/kvm/virtio.h: In function ‘virt_queue__available’:
> >include/kvm/virtio.h:42:2: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]
i suspect the contrast might be from casting a 'struct
vring_used_elem's 'id' field to type '__u16 *' and dereferencing
it might break GCC alias optimizations, as it makes two uses of
the 'num' field - one the regular 32-bit usage, the other this
weird 16-bit usage.
I think the only sane way to solve this is to do what the kernel
does, to turn off strict aliasing. The patch below does this and
resolves the build bug. Note: i also switched optimization from
-Os to -O2 - the latter is generally the better option for
performance critical code. -Os sometimes produces really weird
code.
The other build problem is that it appears the default GCC
regparm model changed, which highlighted this prototype bug:
x86/bios/e820.c:32:15: error: conflicting types for ‘e820_query_map’
In file included from x86/bios/e820.c:1:0:
include/kvm/e820.h:10:6: note: previous declaration of ‘e820_query_map’ was here
and there are similar problems with other BIOS prototypes.
Resolved via the other bits in the patch below.
Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Asias He [Mon, 9 Jan 2012 09:18:01 +0000 (17:18 +0800)]
kvm tools: Fix kvm-ipc.c build breakage
CC kvm-ipc.o
kvm-ipc.c: In function ???kvm_ipc__send_msg???:
kvm-ipc.c:53:34: error: comparison between signed and unsigned integer
expressions [-Werror=sign-compare]
cc1: all warnings being treated as errors
Reported-by: Jean-Philippe Menil <jean-philippe.menil@univ-nantes.fr> Signed-off-by: Asias He <asias.hejun@gmail.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Hongyong Zang [Fri, 6 Jan 2012 02:06:26 +0000 (10:06 +0800)]
kvm tool: Change kvm->ram_size to real mapped size.
If a guest's ram_size exceeds KVM_32BIT_GAP_START, the corresponding kvm tool's
virtual address size should be (ram_size + KVM_32BIT_GAP_SIZE), rather than ram_size.
Use macro define KVM_32BIT_MAX_MEM_SIZE instead of magic number "0x100000000ULL".
Signed-off-by: Hongyong Zang <zanghongyong@huawei.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Pekka Enberg [Tue, 3 Jan 2012 17:49:49 +0000 (19:49 +0200)]
kvm tools, 9p: Fix memory leak in virtio_9p_attach()
This patch fixes the following valgrind reported memory leak:
==7684== 20 bytes in 8 blocks are definitely lost in loss record 3 of 46
==7684== at 0x4A0649D: malloc (vg_replace_malloc.c:236)
==7684== by 0x4116B5: virtio_p9_pdu_readf (9p-pdu.c:110)
==7684== by 0x40FD8C: virtio_p9_attach (9p.c:354)
==7684== by 0x410E80: virtio_p9_do_io (9p.c:1154)
==7684== by 0x40DB45: thread_pool__threadfunc (threadpool.c:56)
==7684== by 0x3DE7007B30: start_thread (in /lib64/libpthread-2.14.so)
==7684== by 0x3DE68DFD2C: clone (in /lib64/libc-2.14.so)
Pekka Enberg [Tue, 3 Jan 2012 17:39:14 +0000 (19:39 +0200)]
kvm tools, 9p: Fix memory leak in virtio_p9_walk()
This patch fixes the following valgrind reported memory leak in
virtio_9p_walk():
==6689== 1,058 bytes in 134 blocks are definitely lost in loss record 39 of 47
==6689== at 0x4A0649D: malloc (vg_replace_malloc.c:236)
==6689== by 0x4116A9: virtio_p9_pdu_readf (9p-pdu.c:110)
==6689== by 0x40FB9F: virtio_p9_walk (9p.c:305)
==6689== by 0x410E73: virtio_p9_do_io (9p.c:1151)
==6689== by 0x40DB45: thread_pool__threadfunc (threadpool.c:56)
==6689== by 0x3DE7007B30: start_thread (in /lib64/libpthread-2.14.so)
==6689== by 0x3DE68DFD2C: clone (in /lib64/libc-2.14.so)
Pekka Enberg [Tue, 3 Jan 2012 17:25:28 +0000 (19:25 +0200)]
kvm tools: Free vcpu array on shutdown
This patch silences the following harmless warning to reduce valgrind noise:
==5529== 864 bytes in 3 blocks are possibly lost in loss record 39 of 48
==5529== at 0x4A04B84: calloc (vg_replace_malloc.c:467)
==5529== by 0x3DE6410D44: _dl_allocate_tls (in /lib64/ld-2.14.so)
==5529== by 0x3DE70082F5: pthread_create@@GLIBC_2.2.5 (in /lib64/libpthread-2.14.so)
==5529== by 0x405755: kvm_cmd_run (builtin-run.c:1169)
==5529== by 0x40D47D: handle_command (kvm-cmd.c:84)
==5529== by 0x3DE682139C: (below main) (in /lib64/libc-2.14.so)
Pekka Enberg [Tue, 3 Jan 2012 17:20:23 +0000 (19:20 +0200)]
kvm tools: Fix use after free during shutdown
Valgrind reports the following use after free error, when shutting down a
guest:
[penberg@tux kvm]$ valgrind ./vm run
[ snip ]
Mounting...
Starting '/bin/sh'...
sh-4.2# exit
[ snip ]
==4726== Thread 11:
==4726== Invalid read of size 8
==4726== at 0x407818: kvm__continue (kvm.c:518)
==4726== by 0x3DE683534F: ??? (in /lib64/libc-2.14.so)
==4726== by 0x3DE68D8AF6: ioctl (in /lib64/libc-2.14.so)
==4726== by 0x406C95: kvm_cpu__run (kvm-cpu.c:35)
==4726== by 0x406D95: kvm_cpu__start (kvm-cpu.c:109)
==4726== by 0x4046D5: kvm_cpu_thread (builtin-run.c:603)
==4726== by 0x3DE7007B30: start_thread (in /lib64/libpthread-2.14.so)
==4726== by 0x3DE68DFD2C: clone (in /lib64/libc-2.14.so)
==4726== Address 0x4c2e040 is 0 bytes inside a block of size 936 free'd
==4726== at 0x4A055FE: free (vg_replace_malloc.c:366)
==4726== by 0x4046E7: kvm_cpu_thread (builtin-run.c:606)
==4726== by 0x3DE7007B30: start_thread (in /lib64/libpthread-2.14.so)
==4726== by 0x3DE68DFD2C: clone (in /lib64/libc-2.14.so)
==4726==
==4726== Thread 1:
==4726== Invalid read of size 1
==4726== at 0x4057BC: kvm_cmd_run (builtin-run.c:1182)
==4726== by 0x40D46D: handle_command (kvm-cmd.c:84)
==4726== by 0x3DE682139C: (below main) (in /lib64/libc-2.14.so)
==4726== Address 0x4e36f58 is 920 bytes inside a block of size 936 free'd
==4726== at 0x4A055FE: free (vg_replace_malloc.c:366)
==4726== by 0x4046E7: kvm_cpu_thread (builtin-run.c:606)
==4726== by 0x3DE7007B30: start_thread (in /lib64/libpthread-2.14.so)
==4726== by 0x3DE68DFD2C: clone (in /lib64/libc-2.14.so)
This patch fixes the problem by moving the kvm_cpu__delete() call from
kvm_cpu_thread() to kvm_cmd_run() after we've made sure the VCPU thread is
stopped.
Lai Jiangshan [Tue, 27 Dec 2011 07:36:38 +0000 (15:36 +0800)]
kvm tools: Don't remove any file which is not created by itself
It is user's or other program's reponsibility, not the current program.
Just try best to let the user know what happen.
Also make print_guest() keep the same behavior as do_debug()
do_pause() ...etc. when it failed.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Thomas Gleixner [Wed, 21 Dec 2011 15:38:10 +0000 (16:38 +0100)]
kvm tool: Implement a serial FIFO emulation
Adding FIFO functionality to the serial emulator to speed up the
console handling.
This also restricts sysrq functionality to the first serial port and
fixes the sysrq injection, which was broken because nothing ever reset
the BI status bit in the LSR.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Asias He [Wed, 21 Dec 2011 15:11:10 +0000 (23:11 +0800)]
kvm tools: Fix compat message per device for virtio/blk
The commit 2b6c246215e8f936d2366ef8c4a6eec730b819b2
(kvm tools: Use compat message per device instead of per instance)
prevents multiple messages for the same type of device.
The variable compat_id is expected to be initialized to -1.
Signed-off-by: Asias He <asias.hejun@gmail.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Lai Jiangshan [Tue, 20 Dec 2011 09:08:51 +0000 (17:08 +0800)]
kvm tools: add kvm_ipc__send() and kvm_ipc__send_msg()
Current code write the sock manually. There is nothing constrains the format of
the written data is expect. Use kvm_ipc__send() and kvm_ipc__send_msg() for
such constraint.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Cyrill Gorcunov [Sun, 18 Dec 2011 20:24:56 +0000 (00:24 +0400)]
kvm tools: Make kvm__arch_setup_firmware to return error code
If some of subsequent calls fails we better to return error
code instead of dying with a message. This is a first step
in getting rid of number of die() calls we have in code.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Cyrill Gorcunov [Sun, 18 Dec 2011 15:16:09 +0000 (19:16 +0400)]
kvm tools: Define __compiletime_error helper
To eliminate compile errors like
| CC builtin-run.o
| In file included from ../../arch/x86/include/asm/system.h:7:0,
| from include/kvm/barrier.h:13,
| from builtin-run.c:16:
| ../../arch/x86/include/asm/cmpxchg.h:11:13: error: no previous prototype for ‘__xchg_wrong_size’ [-Werror=missing-prototypes]
| ../../arch/x86/include/asm/cmpxchg.h: In function ‘__xchg_wrong_size’:
| ../../arch/x86/include/asm/cmpxchg.h:12:2: error: expected declaration specifiers before ‘__compiletime_error’
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Wanlong Gao [Sat, 17 Dec 2011 12:17:19 +0000 (20:17 +0800)]
kvm tools: prevent the redefined error
In file included from ../../include/linux/stddef.h:4:0,
from ../../include/linux/list.h:5,
from include/kvm/framebuffer.h:5,
from builtin-run.c:8:
include/linux/compiler.h:8:0: error: "__attribute_const__" redefined [-Werror]
/usr/include/sys/cdefs.h:226:0: note: this is the location of the previous definition
cc1: all warnings being treated as errors
make: *** [builtin-run.o] Error 1
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Lai Jiangshan [Fri, 16 Dec 2011 09:12:08 +0000 (17:12 +0800)]
kvm-tools: fix wrong "./lkvm list -i"
because the program name is changed, "./lkvm list -i" can't work now.
This fix just remove the test:
1) this test is not reliable.
2) the target process can respond for "get_pid()", it must be
kvm-tool. (at least it is more reliable)
Acked-by: Sasha Levin <levinsasha928@gmail.com> Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Lai Jiangshan [Fri, 16 Dec 2011 09:12:05 +0000 (17:12 +0800)]
kvm-tools: write the sandbox command exactly as expect
sandbox.sh is a shell script, it will translate its meta
chars to anather things.
example(before this patch applied):
./lkvm sandbox -- echo 'success; please try again!'
it fails with:
/virt/sandbox.sh: line 3: please: command not found
Acked-by: Sasha Levin <levinsasha928@gmail.com> Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Matt Evans [Thu, 15 Dec 2011 07:13:41 +0000 (18:13 +1100)]
kvm tools: Fix running with --initrd and no real discs/rootfs
It's useful to boot with no -d and no implicit 9p root, with --initrd.
Since commit eaf720b285947a6f4e29174d0eba1899de31d8ab, we die() failing
to create the 2nd stage custom rootfs. This patch doesn't try to set up
a custom rootfs by default if user specifies an initrd on the commandline.
Signed-off-by: Matt Evans <matt@ozlabs.org> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Matt Evans [Thu, 8 Dec 2011 06:57:09 +0000 (17:57 +1100)]
PPC: Add __SANE_USERSPACE_TYPES__ to asm/types.h for LL64
PPC64 uses long long for u64 in the kernel, but powerpc's asm/types.h
prevents 64-bit userland from seeing this definition, instead defaulting
to u64 == long in userspace. Some user programs (e.g. kvmtool) may actually
want LL64, so this patch adds a check for __SANE_USERSPACE_TYPES__ so that,
if defined, int-ll64.h is included instead.
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Matt Evans <matt@ozlabs.org> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Matt Evans [Wed, 14 Dec 2011 02:06:44 +0000 (13:06 +1100)]
kvm tools: Fix alaising errors in LINT assignment
I'm getting this on x86-32, gcc 4.4.3:
CC x86/kvm-cpu.o
cc1: warnings being treated as errors
x86/kvm-cpu.c: In function ‘kvm_cpu__set_lint’:
x86/kvm-cpu.c:89: error: dereferencing pointer ‘lapic’ does break strict-aliasing rules
x86/kvm-cpu.c:88: error: dereferencing pointer ‘lapic’ does break strict-aliasing rules
x86/kvm-cpu.c:83: note: initialized from here
make: *** [x86/kvm-cpu.o] Error 1
Removing the nasty aliasing seems to be a good way to go.
Acked-by: Sasha Levin <levinsasha928@gmail.com> Signed-off-by: Matt Evans <matt@ozlabs.org> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Matt Evans [Thu, 15 Dec 2011 05:06:57 +0000 (16:06 +1100)]
kvm tools: Remove KVM_NR_CPUS
The KVM_NR_CPUS define is only really used to statically size the global
kvm_cpus array, which can just as easily be allocated on startup. There is
some checking of the -c <nr cpus> value given against NR_CPUs but this is
later again checked against a dynamically-determined limit from
KVM_CAP_MAX_VCPUS anyway. The hardwired limit is arbitrary and not strictly
necessary.
This patch removes the #define, replacing the statically-sized array with
a malloc; the array is kvm->nrcpus+1 in size so that any iterator can halt
at the end (this is done in kvm_cpu__reboot, which doesn't have access to
a struct kvm* and therefore kvm->nrcpus).
An unused #define in x86/mptable.c is also removed.
Signed-off-by: Matt Evans <matt@ozlabs.org> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Sasha Levin [Wed, 14 Dec 2011 13:11:21 +0000 (15:11 +0200)]
kvm tools: Don't always coalesce MMIOs
Coalescing MMIOs is only required when we really want avoid having to exit
to userspace for every MMIO write, usually it's not the case so no reason
to coalesce every MMIO zone (I was also pretty sure it didn't always coalesce
zones)..
Signed-off-by: Sasha Levin <levinsasha928@gmail.com> Signed-off-by: Pekka Enberg <penberg@kernel.org>
Matt Evans [Tue, 13 Dec 2011 07:00:26 +0000 (18:00 +1100)]
kvm tools: Add initial SPAPR PPC64 architecture support
This patch adds a new arch directory, powerpc, basic file structure, register
setup and where necessary stubs out arch-specific functions (e.g. interrupts,
runloop exits) that later patches will provide. The target is an
SPAPR-compliant PPC64 machine (i.e. pSeries); there is no support for PPC32 or
'bare metal' PPC64 guests as yet. Subsequent patches implement the hcalls and
RTAS required to boot SPAPR pSeries kernels.
Memory is mapped from hugetlbfs (as that is currently required by upstream PPC64
HV-mode KVM). The mapping of a VRMA region is yet to be implemented; this is
only necessary on processors that don't support VRMA, e.g. <= P6. Work is
therefore needed to get this going on pre-P7 CPUs.
Processor state is set up as a guest kernel would expect (both primary and
secondaries), and SMP is fully supported.
Finally, support is added for simply loading flat binary kernels (plus initrd).
(bzImages are not used on PPC, and this series does not add zImage support or an
ELF loader.) The intention is to later support loading firmware such as SLOF.
Signed-off-by: Matt Evans <matt@ozlabs.org> Signed-off-by: Pekka Enberg <penberg@kernel.org>
projects / karo-tx-linux.git / log