On Sun, Aug 23, 2015 at 18:14:58 -0700, Paolo Bonzini wrote:
> On 23/08/2015 17:23, Emilio G. Cota wrote:
> > This paves the way for a lockless tb_find_fast.
> >
> > Signed-off-by: Emilio G. Cota
> > ---
(snip)
> > @@ -1707,12 +1735,14 @@ void tb_flush_jmp_
On Sun, Aug 23, 2015 at 18:07:04 -0700, Paolo Bonzini wrote:
> On 23/08/2015 17:24, Emilio G. Cota wrote:
> > Note that user-only uses mmap_lock for this.
> >
> > Signed-off-by: Emilio G. Cota
>
> Why is this needed? The RCU-like page_find should work just fine.
On Sun, Aug 23, 2015 at 18:24:55 -0700, Paolo Bonzini wrote:
> On 23/08/2015 17:24, Emilio G. Cota wrote:
> > This is similar in intent to the async_safe_work mechanism. The main
> > differences are:
> >
> > - Work is run on a single CPU thread *after*
On Sun, Aug 23, 2015 at 18:29:33 -0700, Paolo Bonzini wrote:
>
>
> On 23/08/2015 17:24, Emilio G. Cota wrote:
> > Signed-off-by: Emilio G. Cota
> > ---
> > cputlb.c | 41 +++--
> > 1 file changed, 11 insertions(+), 30 deletio
On Sun, Aug 23, 2015 at 19:01:28 -0700, Paolo Bonzini wrote:
> > * tb_flush: do it once all other CPUs have been put to sleep by calling
> > rcu_synchronize().
> > We also instrument tb_lock to make sure that only one tb_flush request
> > can
> > happen at a given time.
>
> What do
On Wed, Aug 26, 2015 at 14:18:24 +0200, Frederic Konrad wrote:
> Do that make sense?
A few decisions here don't make that much sense to me, but maybe
I'm missing context:
> I'm trying to do the next version of the MTTCG work:
>
> I would like to rebase on Alvise atomic instruction branch:
> -
On Wed, Aug 26, 2015 at 02:17:42 +0200, Paolo Bonzini wrote:
> Signed-off-by: Paolo Bonzini
> ---
> cpu-exec.c | 6 +-
> qom/cpu.c | 2 ++
> 2 files changed, 7 insertions(+), 1 deletion(-)
I like this patch.
Are we making sure that other writes to tcg_exit_req are preceded by
a write barri
On Tue, Sep 01, 2015 at 17:10:30 +0100, Alex Bennée wrote:
>
> Emilio G. Cota writes:
>
> > Signed-off-by: Emilio G. Cota
> > ---
> > cputlb.c | 41 +++--
> > 1 file changed, 11 insertions(+), 30 deletions(-)
>
> I
Hi,
I'm having trouble running a simple multithreaded program on a PowerPC host
machine.
The machine I'm using is a ppc VM--I think it's running under KVM (I'm using
OVH's RunAbove Power8 service):
admin@adsf:~/qemu$ uname -a
Linux adsf 3.13.0-37-generic #64-Ubuntu SMP Mon Sep 22 21:27:09 UT
On Wed, Jun 17, 2015 at 09:58:27 +0100, Peter Maydell wrote:
> On 17 June 2015 at 01:52, Emilio G. Cota wrote:
> > I'm having trouble running a simple multithreaded program on a PowerPC host
> > machine.
> >
> > The machine I'm using is a ppc VM--I th
_phys_page_range->tb_gen_code. I'm not using
gdb so I guess I cannot trigger this.
Am I missing something?
> On 17 June 2015 at 22:36, Emilio G. Cota wrote:
> > I don't think this is a race because it also breaks when
> > run on a single core (with taskset
I cannot make dotprod_mutex.c to crash with the current master (git
8ffe756d). I've tried both linux-arm and linux-arm-static, the latter
running under chroot.
I've tried on three different machines, and have tested with different thread
counts: 4, 10, 16, 64 (one of the machines has 64 cores).
I
lity that
> thread A calls tb_gen_code, which calls tb_alloc, which
> calls tb_flush, which clears the whole code cache, and then
> tb_gen_code starts generating code over the top of a TB
> that thread B was in the middle of executing from...
Agreed, this needs to be fixed. Certainly no
Since 5e5f07e08 "TCG: Move translation block variables
to new context inside tcg_ctx: tb_ctx" on Feb 1 2013, compilation
of usermode + TB_DEBUG_CHECK has been broken. Fix it.
Signed-off-by: Emilio G. Cota
---
translate-all.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
di
e.g. most bootup code is
thrown away after boot); it makes sense to grow the hash table as
more code blocks are translated. This also avoids the complication of
having to build downsizing hysteresis logic into qht.
Signed-off-by: Emilio G. Cota
---
cpu-exe
I'm assuming windows compilers don't support this attribute.
Signed-off-by: Emilio G. Cota
---
include/qemu/compiler.h | 10 ++
1 file changed, 10 insertions(+)
diff --git a/include/qemu/compiler.h b/include/qemu/compiler.h
index 8f1cc7b..fb946f1 100644
--- a/include/qemu/
Signed-off-by: Emilio G. Cota
---
tests/.gitignore | 1 +
tests/Makefile | 6 +++-
tests/test-qht.c | 100 +++
3 files changed, 106 insertions(+), 1 deletion(-)
create mode 100644 tests/test-qht.c
diff --git a/tests/.gitignore b/tests
It is a more appropriate name, now that the mutex embedded
in the seqlock is gone.
Signed-off-by: Emilio G. Cota
---
cpus.c | 28 ++--
include/qemu/seqlock.h | 4 ++--
2 files changed, 16 insertions(+), 16 deletions(-)
diff --git a/cpus.c b/cpus.c
index
or other
workloads that do not translate as many blocks (600K+ for debian-jessie arm
bootup). This is dealt with later in this series.
Signed-off-by: Emilio G. Cota
---
cpu-exec.c | 2 +-
include/exec/tb-hash.h | 19 +--
translate-all.c| 6 +++---
3 file
This option is unused; besides, it bloats the struct when not needed.
Let's just let writers define their own locks elsewhere.
Signed-off-by: Emilio G. Cota
---
cpus.c | 2 +-
include/qemu/seqlock.h | 10 +-
2 files changed, 2 insertions(+), 10 deletions(-)
diff
Wrap pthread_spin on POSIX, or QemuMutex on Windows.
AFAIK there are is no off-the-shelf spinlock implementation for
Windows, so we'll just use QemuMutex.
Signed-off-by: Emilio G. Cota
---
include/qemu/spinlock-posix.h | 60 +++
include/qemu/spi
as a 64-bit implementation, can be found at:
https://github.com/Cyan4973/xxHash
Signed-off-by: Emilio G. Cota
---
include/qemu/xxhash.h | 106 ++
1 file changed, 106 insertions(+)
create mode 100644 include/qemu/xxhash.h
diff --git a/include
we can already
benefit from the single-threaded speedup that qht also provides.
Signed-off-by: Emilio G. Cota
---
include/qemu/qht.h | 150 ++
util/Makefile.objs | 2 +-
util/qht.c | 458 +
3 files changed, 609
This patchset is derived from my ongoing work on MTTCG, but does
not depend on it and brings improvements that we can already
benefit from. It applies cleanly on the current master and
is checkpatch-clean.
The key goal is to make the TB hash table faster, and while at it,
scalable. Tested on two d
On Tue, Apr 05, 2016 at 08:57:45 +0100, Peter Maydell wrote:
> On 5 April 2016 at 06:30, Emilio G. Cota wrote:
> > +#define QEMU_CACHELINE (64)
>
> Why 64? Does anything bad happen if the host's cache line
> size turns out to be greater than the value here ?
Defining
On Tue, Apr 05, 2016 at 19:01:07 +0100, Peter Maydell wrote:
> On 5 April 2016 at 18:24, Emilio G. Cota wrote:
> > So how about this:
> > we add these defaults, and also add an optional --configure
> > parameter to override said defaults.
>
> I think this definitel
On Tue, Apr 05, 2016 at 09:07:57 -0700, Richard Henderson wrote:
> On 04/05/2016 08:48 AM, Paolo Bonzini wrote:
> >I think it's fine to use the struct. The exact size of the struct
> >varies from 3 to 5 32-bit words, so it's hard to write nice
> >size-dependent code for the hash.
>
> I don't thin
look?
Thanks,
E.
commit af92a0690f49172621cd8b80759e3ca567d43567
Author: Emilio G. Cota
Date: Tue Apr 5 18:06:21 2016 -0400
rth
Signed-off-by: Emilio G. Cota
diff --git a/include/exec/tb-hash.h b/include/exec/tb-hash.h
index 6b97a7c..349a856 100644
--- a/i
boots OK for me with the patch below. I'm booting
it as per the instructions in
http://www.bennee.com/~alex/blog/2014/05/09/running-linux-in-qemus-aarch64-system-emulation-mode/
Thanks,
Emilio
commit e70474788fa37a85df21e1c63101a879103758f5
Author: Emilio G. Cota
Date:
On Wed, Apr 06, 2016 at 13:52:21 +0200, Paolo Bonzini wrote:
>
>
> On 06/04/2016 02:52, Emilio G. Cota wrote:
> > +static inline uint32_t tb_hash_func5(uint64_t a0, uint64_t b0, uint32_t e,
> > int seed)
>
> I would keep just this version and unconditionally zer
On Wed, Mar 23, 2016 at 16:12:46 +, Peter Maydell wrote:
> On 17 March 2016 at 13:25, Shannon Zhao wrote:
> > From: Shannon Zhao
> >
> > There is a problem for power button that it will not work if an early
> > system_powerdown request happens before guest gpio driver loads.
> >
> > Here we a
On Wed, Apr 06, 2016 at 12:39:42 +0100, Alex Bennée wrote:
>
> Emilio G. Cota writes:
>
> > xxhash is a fast, high-quality hashing function. The appended
> > brings in the 32-bit version of it, with the small modification that
> > it assumes the data to be hashed is
On Wed, Apr 06, 2016 at 20:23:42 +0200, Paolo Bonzini wrote:
> On 06/04/2016 19:44, Emilio G. Cota wrote:
> > I like this idea, because the ugliness of the sizeof checks is significant.
> > However, the quality of the resulting hash is not as good when always using
> > fu
Taken from the linux kernel.
Signed-off-by: Emilio G. Cota
---
include/qemu/processor.h | 28
1 file changed, 28 insertions(+)
create mode 100644 include/qemu/processor.h
diff --git a/include/qemu/processor.h b/include/qemu/processor.h
new file mode 100644
index
-by: Laurent Desnogues
Suggested-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
cpu-exec.c | 6 +++---
exec.c | 2 +-
hw/i386/kvmvapic.c | 2 +-
include/exec/exec-all.h | 5 +++--
target-alpha/cpu.h | 2 +-
target-arm/cpu.h| 2 +-
tar
Signed-off-by: Emilio G. Cota
---
tests/.gitignore | 1 +
tests/Makefile | 6 ++-
tests/test-qht.c | 109 +++
3 files changed, 115 insertions(+), 1 deletion(-)
create mode 100644 tests/test-qht.c
diff --git a/tests/.gitignore b/tests
This will be used by upcoming changes for hashing the tb hash.
Add this into a separate file to include the copyright notice from
xxhash.
Signed-off-by: Emilio G. Cota
---
include/exec/tb-hash-xx.h | 103 ++
1 file changed, 103 insertions(+)
create
It is a more appropriate name, now that the mutex embedded
in the seqlock is gone.
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
cpus.c | 28 ++--
include/qemu/seqlock.h | 4 ++--
2 files changed, 16 insertions(+), 16 deletions(-)
diff
From: Guillaume Delbergue
Signed-off-by: Guillaume Delbergue
[Rewritten. - Paolo]
Signed-off-by: Paolo Bonzini
---
include/qemu/thread.h | 31 +++
1 file changed, 31 insertions(+)
diff --git a/include/qemu/thread.h b/include/qemu/thread.h
index bdae6df..1aa843b 100
or other
workloads that do not translate as many blocks (600K+ for debian-jessie arm
bootup). This is dealt with later in this series.
Signed-off-by: Emilio G. Cota
---
cpu-exec.c | 2 +-
include/exec/tb-hash.h | 8 ++--
translate-all.c| 6 +++---
3 files changed, 10 inser
Signed-off-by: Emilio G. Cota
---
include/qemu/compiler.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/qemu/compiler.h b/include/qemu/compiler.h
index 8f1cc7b..1978ddc 100644
--- a/include/qemu/compiler.h
+++ b/include/qemu/compiler.h
@@ -41,6 +41,8 @@
# define QEMU_PACKED
je 4ad980
4ad996: f3 90 pause
4ad998: eb f6 jmp4ad990
Signed-off-by: Emilio G. Cota
---
include/qemu/thread.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/include/qemu/thread.h b/include/qemu/thread.h
Signed-off-by: Emilio G. Cota
---
include/qemu/thread.h | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/include/qemu/thread.h b/include/qemu/thread.h
index 1aa843b..599965e 100644
--- a/include/qemu/thread.h
+++ b/include/qemu/thread.h
@@ -2,6 +2,7 @@
#define
On Thu, Apr 07, 2016 at 13:19:22 -0400, Emilio G. Cota wrote:
> We are inconsistent with the type of tb->flags: usage varies loosely
> between int and uint64_t. Settle to uint32_t everywhere, which is
> superior to both: at least one target (aarch64) uses the most significant
> bit
See v1 for context:
https://lists.gnu.org/archive/html/qemu-devel/2016-04/msg00587.html
All patches in v2 are checkpatch-clean, except 05 (checkpatch should
be ignored for this one) and 06, which I took unmodified (later patches
fix those warnings while doing other things, anyway).
Note that pat
This option is unused; besides, it bloats the struct when not needed.
Let's just let writers define their own locks elsewhere.
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
cpus.c | 2 +-
include/qemu/seqlock.h | 10 +-
2 files changed, 2 insertions(+
On Fri, Apr 08, 2016 at 11:45:41 +0100, Alex Bennée wrote:
(snip the entire patch)
> A couple of notes:
>
> - these should use the gtester boiler plate for reporting results
Done in v3.
> - AFAICT they are not exercising the multi-element hashing we actually
> use in the main code
> -
Hi Alex,
I'm sending a v3 in a few minutes. I've addressed all your comments there,
so I won't duplicate them here; please find inline my replies to some
questions you raised.
On Fri, Apr 08, 2016 at 11:27:19 +0100, Alex Bennée wrote:
> Emilio G. Cota writes:
(snip)
> >
This option is unused; besides, it bloats the struct when not needed.
Let's just let writers define their own locks elsewhere.
Reviewed-by: Alex Bennée
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
cpus.c | 2 +-
include/qemu/seqlock.h | 10 +---
It is a more appropriate name, now that the mutex embedded
in the seqlock is gone.
Reviewed-by: Alex Bennée
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
cpus.c | 28 ++--
include/qemu/seqlock.h | 4 ++--
2 files changed, 16
e.g. most bootup code is
thrown away after boot); it makes sense to grow the hash table as
more code blocks are translated. This also avoids the complication of
having to build downsizing hysteresis logic into qht.
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
cpu-ex
From: Guillaume Delbergue
Signed-off-by: Guillaume Delbergue
[Rewritten. - Paolo]
Signed-off-by: Paolo Bonzini
[Emilio's additions: call cpu_relax() while spinning; optimize for
uncontended locks by acquiring the lock with xchg+test instead of
test+xchg+test.]
Signed-off-by: Emilio G.
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
include/qemu/compiler.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/qemu/compiler.h b/include/qemu/compiler.h
index 8f1cc7b..1978ddc 100644
--- a/include/qemu/compiler.h
+++ b/include/qemu/compiler.h
@@ -41,6
Suggested-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
translate-all.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/translate-all.c b/translate-all.c
index 617a572..769bffc 100644
--- a/translate-all.c
+++ b/translate-all.c
@@ -1664,6 +1664,8 @@ void dump_exec_info(FILE
Signed-off-by: Emilio G. Cota
---
tests/.gitignore | 1 +
tests/Makefile | 6 ++-
tests/test-qht.c | 139 +++
3 files changed, 145 insertions(+), 1 deletion(-)
create mode 100644 tests/test-qht.c
diff --git a/tests/.gitignore b/tests
we can already
benefit from the single-threaded speedup that qht also provides.
Signed-off-by: Emilio G. Cota
---
include/qemu/qht.h | 54 +
util/Makefile.objs | 2 +-
util/qht.c | 590 +
3 files changed, 645 insertions(+), 1
Taken from the linux kernel.
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
include/qemu/processor.h | 28
1 file changed, 28 insertions(+)
create mode 100644 include/qemu/processor.h
diff --git a/include/qemu/processor.h b/include/qemu
See v2 here:
https://lists.gnu.org/archive/html/qemu-devel/2016-04/msg01307.html
Changes from v2:
- Dropped "add missing fold of tb_ctx into tcg_ctx", already merged
upstream as commit 7e6bd36d611.
- Added reviewed-by tags from Alex and Richard
- xxhash:
+ use rol32 from qemu/bitops.h
+
This will be used by upcoming changes for hashing the tb hash.
Add this into a separate file to include the copyright notice from
xxhash.
Signed-off-by: Emilio G. Cota
---
include/exec/tb-hash-xx.h | 94 +++
1 file changed, 94 insertions(+)
create
or other
workloads that do not translate as many blocks (600K+ for debian-jessie arm
bootup). This is dealt with later in this series.
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
cpu-exec.c | 2 +-
include/exec/tb-hash.h | 8 ++--
translate-all.c|
On Thu, Apr 07, 2016 at 13:19:22 -0400, Emilio G. Cota wrote:
> We are inconsistent with the type of tb->flags: usage varies loosely
> between int and uint64_t. Settle to uint32_t everywhere, which is
> superior to both: at least one target (aarch64) uses the most significant
> bit
On Wed, Apr 20, 2016 at 14:15:41 +0200, KONRAD Frederic wrote:
> Le 20/04/2016 01:07, Emilio G. Cota a écrit :
> >Taken from the linux kernel.
> >
> >Reviewed-by: Richard Henderson
> >Signed-off-by: Emilio G. Cota
> >---
> > include/qemu/processor.h | 28 +
On Wed, Apr 20, 2016 at 08:18:56 -0700, Richard Henderson wrote:
> On 04/19/2016 04:07 PM, Emilio G. Cota wrote:
> >From: Guillaume Delbergue
> >
> >Signed-off-by: Guillaume Delbergue
> >[Rewritten. - Paolo]
> >Signed-off-by: Paolo Bonzini
> >[Emilio'
Taken from the linux kernel.
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
include/qemu/processor.h | 34 ++
1 file changed, 34 insertions(+)
create mode 100644 include/qemu/processor.h
diff --git a/include/qemu/processor.h b/include/qemu
milio G. Cota
---
include/qemu/thread.h | 34 ++
1 file changed, 34 insertions(+)
diff --git a/include/qemu/thread.h b/include/qemu/thread.h
index bdae6df..a216941 100644
--- a/include/qemu/thread.h
+++ b/include/qemu/thread.h
@@ -1,6 +1,9 @@
#ifndef __QEMU_THR
On Wed, Apr 20, 2016 at 10:55:45 -0700, Richard Henderson wrote:
> On 04/20/2016 10:17 AM, Emilio G. Cota wrote:
> >I've tried to find a GCC intrinsic for test-and-set, and I've only found
> >lock_test_and_set, which is what we use for atomic_xchg (except on ppc)
> >
On Wed, Apr 20, 2016 at 12:39:45 -0700, Richard Henderson wrote:
> On 04/20/2016 11:11 AM, Emilio G. Cota wrote:
> >On Wed, Apr 20, 2016 at 10:55:45 -0700, Richard Henderson wrote:
> >>On 04/20/2016 10:17 AM, Emilio G. Cota wrote:
(snip)
> >My comment was related to thi
Signed-off-by: Emilio G. Cota
---
translate-all.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/translate-all.c b/translate-all.c
index e700399..bba9b62 100644
--- a/translate-all.c
+++ b/translate-all.c
@@ -668,39 +668,39 @@ static inline void *alloc_code_gen_buffer
=
size_code_gen_buffer(tb_size);
Signed-off-by: Emilio G. Cota
---
translate-all.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/translate-all.c b/translate-all.c
index 769bffc..e700399 100644
--- a/translate-all.c
+++ b/translate-all.c
@@ -505,7 +505,6 @@ static inline size_t size_code_gen_buffer
g the assert. If you have a working
multi-threaded workload that would be good to test this, please
let me know.
- Windows; not even compile-tested!
Signed-off-by: Emilio G. Cota
---
translate-all.c | 122 +---
1 file changed, 117 insertions(
On Fri, Apr 22, 2016 at 10:35:42 +0100, Peter Maydell wrote:
> On 20 April 2016 at 00:07, Emilio G. Cota wrote:
> > +#define QEMU_ALIGNED(B) __attribute__((aligned(B)))
>
> A rather trivial thing, but if we have to respin this series
> for some other reason could we use
On Fri, Apr 22, 2016 at 10:41:25 -0700, Richard Henderson wrote:
> On 04/19/2016 04:07 PM, Emilio G. Cota wrote:
> > +ht_avg_len = qht_avg_bucket_chain_length(&tcg_ctx.tb_ctx.htable,
> > &ht_heads);
> > +cpu_fprintf(f, "TB hash avg ch
On Fri, Apr 22, 2016 at 12:59:52 -0700, Richard Henderson wrote:
> FWIW, so that I could get an idea of how the stats change as we improve the
> hashing, I inserted the attachment 1 patch between patches 5 and 6, and with
> attachment 2 attempting to fix the accounting for patches 9 and 10.
For qh
On Fri, Apr 22, 2016 at 15:41:13 +0100, Alex Bennée wrote:
> Emilio G. Cota writes:
(snip)
> > Known issues:
> > - Basically compile-tested only, since I've only run this with
> > single-threaded TCG; I also tried running it with linux-user,
> > but in order t
Alex' unit test with low enough -tb-size, see
https://lists.gnu.org/archive/html/qemu-devel/2016-04/msg03465.html
Seems to work in MTTCG, although I've only tested with tb_lock
always being taken in tb_find_fast.
- Windows; not even compile-tested!
Signed-off-by: Emilio G. Cota
On Sun, Apr 24, 2016 at 13:01:31 -0700, Richard Henderson wrote:
> On 04/19/2016 04:07 PM, Emilio G. Cota wrote:
> >+static void qht_insert__locked(struct qht *ht, struct qht_map *map,
> >+ struct qht_bucket *head, void *p, uint32_t
> >hash)
On Sun, Apr 24, 2016 at 12:46:08 -0700, Richard Henderson wrote:
> On 04/22/2016 04:57 PM, Emilio G. Cota wrote:
> >On Fri, Apr 22, 2016 at 12:59:52 -0700, Richard Henderson wrote:
> >>FWIW, so that I could get an idea of how the stats change as we improve the
> >
On Mon, Apr 25, 2016 at 16:19:59 +0100, Alex Bennée wrote:
>
> Emilio G. Cota writes:
>
> > [ Applies on top of bennee/mttcg/enable-mttcg-for-armv7-v1 after
> > reverting "translate-all: introduces tb_flush_safe". A trivial
> > conflict must be solved afte
: [<0020>]lr : [<40010800>]psr: 6153
sp : 400b45c0 ip : 400b34e8 fp : 40032ca8
r10: r9 : r8 :
r7 : r6 : r5 : r4 :
r3 : r2 : r1 : 00ff r0 :
Flags: nZCv IRQs on FIQs of
On Sun, Apr 24, 2016 at 18:06:51 -0400, Emilio G. Cota wrote:
> On Sun, Apr 24, 2016 at 12:46:08 -0700, Richard Henderson wrote:
> > On 04/22/2016 04:57 PM, Emilio G. Cota wrote:
> > >On Fri, Apr 22, 2016 at 12:59:52 -0700, Richard Henderson wrote:
> > >>FWIW, so t
Signed-off-by: Emilio G. Cota
---
tests/.gitignore | 1 +
tests/Makefile | 6 +-
tests/test-qdist.c | 363 +
3 files changed, 369 insertions(+), 1 deletion(-)
create mode 100644 tests/test-qdist.c
diff --git a/tests/.gitignore b
It is a more appropriate name, now that the mutex embedded
in the seqlock is gone.
Reviewed-by: Alex Bennée
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
cpus.c | 28 ++--
include/qemu/seqlock.h | 4 ++--
2 files changed, 16
This new helper expands to __atomic_test_and_set where available;
otherwise it expands to atomic_xchg.
Signed-off-by: Emilio G. Cota
---
include/qemu/atomic.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/qemu/atomic.h b/include/qemu/atomic.h
index 5bc4d6c..6132dad 100644
--- a
Taken from the linux kernel.
Reviewed-by: Richard Henderson
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
include/qemu/processor.h | 34 ++
1 file changed, 34 insertions(+)
create mode 100644 include/qemu/processor.h
diff --git a/include/qemu
:
[15.0,16.7)|▁▂▅▄█▅|[30.3,32.0]
Suggested-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
translate-all.c | 36
1 file changed, 36 insertions(+)
diff --git a/translate-all.c b/translate-all.c
index 0bf76d7..775ea79 100644
--- a/translate-all.c
+++ b
or other
workloads that do not translate as many blocks (600K+ for debian-jessie arm
bootup). This is dealt with later in this series.
Reviewed-by: Richard Henderson
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
cpu-exec.c | 4 ++--
include/exec/tb-hash.h | 8 ++-
TATAS.]
Signed-off-by: Emilio G. Cota
---
include/qemu/thread.h | 34 ++
1 file changed, 34 insertions(+)
diff --git a/include/qemu/thread.h b/include/qemu/thread.h
index bdae6df..39ff1ac 100644
--- a/include/qemu/thread.h
+++ b/include/qemu/thread.h
@@ -1,6
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
tests/.gitignore | 1 +
tests/Makefile | 5 +-
tests/test-qht.c | 177 +++
3 files changed, 182 insertions(+), 1 deletion(-)
create mode 100644 tests/test-qht.c
diff --git a
dist, 39);
[...]
qdist_inc(&dist, 80);
char *str = qdist_pr(&dist, 9, QDIST_PR_LABELS);
// -> [39.0,43.6)▂▂ █▂ ▂ ▄[75.4,80.0]
g_free(str);
char *str = qdist_pr(&dist, 4, QDIST_PR_LABELS);
// -> [39.0,49.2)▁█▁▁[69.8,80.0]
g_free(str);
Signed-off-by: Emilio G. Cota
---
include/qemu/qdi
This will be used by upcoming changes for hashing the tb hash.
Add this into a separate file to include the copyright notice from
xxhash.
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
include/exec/tb-hash-xx.h | 94 +++
1 file
On Tue, Apr 26, 2016 at 07:32:39 +0100, Alex Bennée wrote:
> Emilio G. Cota writes:
> > With two code_gen "halves", if two tb_flush calls are done in the same
> > RCU read critical section, we're screwed. I added a cpu_exit at the end
> > of tb_flush to try to
Changes from v3:
- added reviewed-by tags from v3. I dropped the review tags from the
'qht' and 'info jit' patches because they have changed quite a bit
from v3.
- qdist: new module to print intuitive histograms, see 'info jit' below.
- qht:
+ bug fix: remove unnecessary requirement of hashe
we can already
benefit from the single-threaded speedup that qht also provides.
Signed-off-by: Emilio G. Cota
---
include/qemu/qht.h | 67 +
util/Makefile.objs | 2 +-
util/qht.c | 722 +
3 files changed, 790 insertions(+), 1
e.g. most bootup code is
thrown away after boot); it makes sense to grow the hash table as
more code blocks are translated. This also avoids the complication of
having to build downsizing hysteresis logic into qht.
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
cpu-ex
This option is unused; besides, it bloats the struct when not needed.
Let's just let writers define their own locks elsewhere.
Reviewed-by: Alex Bennée
Reviewed-by: Richard Henderson
Signed-off-by: Emilio G. Cota
---
cpus.c | 2 +-
include/qemu/seqlock.h | 10 +---
Reviewed-by: Richard Henderson
Reviewed-by: Alex Bennée
Signed-off-by: Emilio G. Cota
---
include/qemu/compiler.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/qemu/compiler.h b/include/qemu/compiler.h
index 8f1cc7b..b64f899 100644
--- a/include/qemu/compiler.h
+++ b/include
On Tue, May 03, 2016 at 08:36:35 +0100, Alex Bennée wrote:
> Emilio G. Cota writes:
> > static TranslationBlock *tb_find_physical(CPUState *cpu,
> >target_ulong pc,
> >
On Tue, May 03, 2016 at 19:24:31 -1000, Richard Henderson wrote:
> On 05/03/2016 07:26 AM, Emilio G. Cota wrote:
> >Ouch, sorry. Won't happen again.
> >
> >Grab the missing pre-requisite patch from:
> > http://patchwork.ozlabs.org/patch/607662/mbox/
>
&
On Wed, May 04, 2016 at 07:22:16 -1000, Richard Henderson wrote:
> On 05/04/2016 05:36 AM, Emilio G. Cota wrote:
> >BTW in the last couple of days I did some more work beyond v4:
> >
> >- Added a benchmark (not a correctness test) to measure parallel
> > performance o
On Mon, May 09, 2016 at 13:21:50 +0200, Paolo Bonzini wrote:
> On 30/04/2016 05:40, Emilio G. Cota wrote:
> >> The tb_flush
> >> > is a fairly rare occurrence its not like its on the critical performance
> >> > path (although of course pathological cases are p
101 - 200 of 2303 matches
Mail list logo
