Hi Nadav,
It's probably a random patch I wrote in a second state. I don't remember
writing it :(
Best regards
Benoît
2018-03-07 18:17 GMT+01:00 Nadav Har'El :
> Ok, I opened https://github.com/cloudius-systems/osv/issues/951
> where I explain what I think is the real bug
Looks good to me.
Hi Nadav could you review this ?
Best regards
Benoīt
On Wed, Mar 08, 2017 at 08:22:00AM +0100, Justin Cinkelj wrote:
> Before patch, netperf netserver.so and netperf.so could not be run in
> the same VM. The netserver reports "error starting alarm timer,
> ret 3 errno 28".
Anyway qperf seems promising to benchmark stuff but I wonder if it's the
right tools since it fork().
I will google a bit more to find something that does not fork().
2017-10-01 16:24 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
>
> On Sun, Oct 1, 2017 at 5:12 PM,
/proc/stat was in the wrong place.
Move it.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
fs/procfs/procfs_vnops.cc | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/fs/procfs/procfs_vnops.cc b/fs/procfs/procfs_vnops.cc
index dc66ff59..437f8302 100644
--- a/fs/
7 at 1:01 PM, Benoît Canet <ben...@scylladb.com> wrote:
>
>>
>> It's what I get as first result when I google iperf so it does look quite
>> popular.
>>
>
> Can you please do a slightly deeper search? :-)
> Does it seem the same developers have abandoned
It's what I get as first result when I google iperf so it does look quite
popular.
2017-09-18 11:51 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
> On Fri, Sep 15, 2017 at 3:58 PM, Benoît Canet <benoit.canet.contrib@gmail.
> com> wrote:
>
>> This one does compile on rece
oh this is nice.
Are we sure of busybox license ?
2017-08-07 11:32 GMT+02:00 Justin Cinkelj :
> Busybox ash shell is more powerful than OSv --runscript option.
> Use ash when fancy shell is required.
>
> While busybox config file includes default options (e.g. nearly all
I mean you use wrk on the host to generate a lot of requests.
2017-08-02 8:36 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
>
> On Wed, Aug 2, 2017 at 9:30 AM, Benoît Canet <ben...@scylladb.com> wrote:
>
>>
>> simply porting the default hello world to OSv and
simply porting the default hello world to OSv and loading it with wrk would
help exibit the issue.
2017-08-02 8:26 GMT+02:00 Nadav Har'El :
>
> On Wed, Aug 2, 2017 at 7:18 AM, Hawx Chen
> wrote:
>
>> Hi Nadav, thank you for your reply.
>> I already
Sorry ignore this I missed your comments in the middle and bottom of your
mail answer.
--
You received this message because you are subscribed to the Google Groups "OSv
Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to
external/ is an old filesystem snapshot
several years old.
This means the openssl and crypto libraries
in it are obsolettes.
Use the libraries available on the host.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
modules/httpserver/Makefile | 7 +--
modules/openssl/usr.manife
external/ is an old filesystem snapshot
several years old.
This means the openssl and crypto libraries
in it are obsolettes.
Get rid of this particular crypto dependencies
by using the host libraries.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
modules/httpserver/Makefil
If OpenSSL version is 1.1 prompt a message explaining
to the user how to install the compat headers.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
modules/lua/Makefile | 2 ++
modules/lua/check-openssl-version | 16
2 files changed, 18 insertions(+)
If OpenSSL version is 1.1 prompt a message explaining
to the user how to install the compat headers.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
modules/lua/Makefile | 2 ++
modules/openssl/Makefile | 3 +++
scripts/check-openssl-version | 16
3
This reverts commit dc447e66dea85d283fa98425837f1c983adc9067.
---
modules/lua/Makefile | 8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/modules/lua/Makefile b/modules/lua/Makefile
index c4c80144..64ea4170 100644
--- a/modules/lua/Makefile
+++ b/modules/lua/Makefile
@@
Using the default openssl-devel package would break
lua sec and start a dependency mess.
Use the compatibility header package so the build
and run will work just fine.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
scripts/setup.py | 18 +-
1 file changed, 9 inse
think of this ?
Best regards
Benoît
2017-07-24 12:11 GMT+02:00 Benoît Canet <ben...@scylladb.com>:
>
>
> 2017-07-24 8:02 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
>
>>
>> On Sat, Jul 22, 2017 at 1:51 PM, Benoît Canet <
>> benoit.canet.cont...@gm
2017-07-24 8:02 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
>
> On Sat, Jul 22, 2017 at 1:51 PM, Benoît Canet <benoit.canet.contrib@gmail.
> com> wrote:
>
>> The LuaRock repository is not up to date.
>>
>
> Hi,
>
> I am far from a Lua or LuaRock ex
LuaSec stopped compiling because of new openssl version.
2017-07-23 13:01 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
>
> On Sat, Jul 22, 2017 at 7:32 PM, Benoît Canet <ben...@scylladb.com> wrote:
>
>>
>> Maybe a better solution than this patch would be to
Maybe a better solution than this patch would be to avoid the mix of host
openssl and
internal openssl so this patch would be useless.
Additionally it would make the default build run as it is on fedora 26.
2017-07-22 12:51 GMT+02:00 Benoît Canet <benoit.canet.cont...@gmail.com>:
> Th
To be complete misc.bin should be updated and
openssl switched to version 1.1 in openssl manifest.
---
scripts/setup.py | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/scripts/setup.py b/scripts/setup.py
index 9bf9de48..0c4bfe31 100755
--- a/scripts/setup.py
+++
The LuaRock repository is not up to date.
Use the github repository.
Signed-of-by: Benoît Canet <ben...@scylladb.com>
---
modules/lua/Makefile | 8 ++--
modules/openssl/usr.manifest | 2 +-
2 files changed, 7 insertions(+), 3 deletions(-)
diff --git a/modules/lua/Makefile b/m
---
scripts/setup.py | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/scripts/setup.py b/scripts/setup.py
index 9bf9de48..0c4bfe31 100755
--- a/scripts/setup.py
+++ b/scripts/setup.py
@@ -84,7 +84,14 @@ class Fedora(object):
ec2_post_install = None
2017-04-26 0:50 GMT+02:00 Nadav Har'El :
> This patch adds the missing support in ramfs for creating and reading
> symbolic links, and then modifies the "bootfs" format (which we use to
> pack and unpack the ram disk files) to also allow symbolic links.
>
> After this patch,
Reviewed-by: Benoît Canet <ben...@scylladb.com>
2017-04-26 0:50 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
> Remove some dead code (an ifdef'ed out static function which was never
> used)
> from fs/vfs/main.cc.
>
> Signed-off-by: Nadav Har'El <n...@scylladb.com
Reviewed-by: Benoît Canet <ben...@scylladb.com>
2017-04-18 14:29 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
> Issue #869 reported that the bug which was supposedly fixed by commit
> 34232de9 has returned when running OSv in a Linux 4.10 host. This patch
> fixes the same
Reviewed-by: Benoît Canet <ben...@scylladb.com>
2017-01-18 10:32 GMT+01:00 Nadav Har'El <n...@scylladb.com>:
> Allow unpinning a thread with pthread_setaffinity_np() and
> sched_setaffinity() by setting the CPU mask to all CPUs.
>
> We still do not support setting the
Reviewed-by: Benoît Canet <ben...@irqsave.net>
On Sun, Jan 15, 2017 at 10:49 PM, Justin Cinkelj <justin.cink...@xlab.si>
wrote:
> NFS mount options can be in format "uid=NUM,gid=NUM".
> In this case, mount-nfs.so requires replacement of "," with
Could you split it in two patches ?
One to add the latest ubuntu and the other to add the jdks ?
On Fri, Jan 13, 2017 at 10:51 PM, 'Sergiy Kibrik' via OSv Development <
osv-dev@googlegroups.com> wrote:
> Signed-off-by: Sergiy Kibrik
> ---
> scripts/setup.py | 17
2017-01-13 14:35 GMT+01:00 Nadav Har'El :
> This patch fixes a deadlock caused by lock order inversion. This deadlock,
> described in issue #817, sometimes hangs the entire VFS when a mount is
> attempted.
>
> Most of the VFS code takes locks on various nodes, and later, in
Reviewed-by: Benoît Canet <ben...@scylladb.com>
2017-01-13 14:35 GMT+01:00 Nadav Har'El <n...@scylladb.com>:
> There's no reason to use std::mutex where we can use Osv's "mutex".
> std::mutex basically wraps OSv's mutex by a Posix mutex and then C++'s
> mutex impl
Reviewed-by: Benoît Canet <ben...@scylladb.com>
On Fri, Jan 13, 2017 at 1:22 PM, 'Sergiy Kibrik' via OSv Development <
osv-dev@googlegroups.com> wrote:
> Signed-off-by: Sergiy Kibrik <sergiy.kib...@globallogic.com>
> ---
> scripts/setup.py | 9 -
> 1 fi
c768520%28v=bts.10%29.aspx
>
> Please, be specific.
>
>
> On Thursday, December 15, 2016 at 9:05:19 AM UTC-8, Benoît CANET wrote:
>>
>>
>> OSV would need driver for HyperV to be written as described in the github
>> issue I
>> gave you.
>>
>&
2016 at 3:57:23 AM UTC-8, Benoît Canet wrote:
>>
>>
>>
>> --
> You received this message because you are subscribed to the Google Groups
> "OSv Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to osv-d
t; On Wednesday, December 14, 2016 at 3:38:20 AM UTC-8, Benoît Canet wrote:
>>
>>
>> --
> You received this message because you are subscribed to the Google Groups
> "OSv Development" group.
> To unsubscribe from this group and stop receiving emails from it, sen
How can we be sure go does not use static tls ?
On Mon, Oct 31, 2016 at 10:35 AM, Nadav Har'El <n...@scylladb.com> wrote:
>
> On Mon, Oct 31, 2016 at 10:35 AM, Benoît Canet <
> ben...@cloudius-systems.com> wrote:
>
>>
>>
>> On Mon, Oct 31, 2016 at 9:30
On Mon, Oct 31, 2016 at 9:30 AM, Nadav Har'El <n...@scylladb.com> wrote:
>
> On Mon, Oct 31, 2016 at 9:48 AM, Benoît Canet <ben...@cloudius-systems.com
> > wrote:
>
>>
>>
>> On Thu, Oct 27, 2016 at 3:04 PM, Nadav Har'El <n...@scylladb.com> wrote:
&g
popfq to restore rflags.
Fixes #798
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
arch/x64/entry.S | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/arch/x64/entry.S b/arch/x64/entry.S
index cd23cec..04d809d 100644
--- a/arch/x64/entry.S
+++ b/arch/x64/e
This is required by go runtime initialization.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 2 +-
core/elf.cc| 12 ++--
include/osv/elf.hh | 4 ++--
3 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/core/app.cc b/core/app.cc
Use the fpu_lock to do the work of saving and restoring
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
arch/x64/entry.S | 2 +-
linux.cc | 4
2 files changed, 5 insertions(+), 1 deletion(-)
diff --git a/arch/x64/entry.S b/arch/x64/entry.S
index dda7358..cd23cec
This will allow go program to get the vdso library address.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 21 ++---
include/osv/app.hh | 7 ++-
2 files changed, 24 insertions(+), 4 deletions(-)
diff --git a/core/app.cc b/core/app.cc
this.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
Makefile| 7 ++-
bootfs.manifest.skel| 1 +
libc/vdso/vdso.c| 18 ++
libc/vdso/vdso.version | 3 +++
usr.manifest.skel | 1 +
usr_nozfs.manifest.skel | 1 +
6 files chang
of in next patches.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 65 --
include/osv/app.hh | 8 +++
2 files changed, 42 insertions(+), 31 deletions(-)
diff --git a/core/app.cc b/core/app.cc
index ad3145f..5
argv preparation will have to load the vdso compatibility
library and set it's address into the auxv part of argv
_before_ loading the actual program .so.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git
Do not delay static_tls init.
Benoît Canet (11):
app: Factorize argc and argv construction
app: Cleanup argv preparation
app: Prepare argv early
elf/app: Delay elf object initialization
elf/app: Pass argc and argv to library initialization
vdso: vdso fast syscall compatibility library
ficial OSv go main()
entry point. Since the go library init already took argc and
argv as parameters GoMain does not need to bother with them
and hence is specific to Go.:
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc | 3 +++
1 file changed, 3 insertions(+)
diff --git
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
linux.cc | 20
1 file changed, 20 insertions(+)
diff --git a/linux.cc b/linux.cc
index e78bb0c..4ab88f3 100644
--- a/linux.cc
+++ b/linux.cc
@@ -24,6 +24,10 @@
#include
#include
#include
+#include
+#i
We want the init of the elf object to use
the tls defined when the new thread is started.
Delay this initialization.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 4 +++-
core/elf.cc| 49 -
include/osv/
of in next patches.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 65 --
include/osv/app.hh | 8 +++
2 files changed, 42 insertions(+), 31 deletions(-)
diff --git a/core/app.cc b/core/app.cc
index ad3145f..5
this.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
Makefile| 7 ++-
bootfs.manifest.skel| 1 +
libc/vdso/vdso.c| 18 ++
libc/vdso/vdso.version | 3 +++
usr.manifest.skel | 1 +
usr_nozfs.manifest.skel | 1 +
6 files chang
This will allow go program to get the vdso library address.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 21 ++---
include/osv/app.hh | 7 ++-
2 files changed, 24 insertions(+), 4 deletions(-)
diff --git a/core/app.cc b/core/app.cc
Fix small argv type error
Add proper Fixes to rflag patch
Benoît Canet (11):
app: Factorize argc and argv construction
app: Cleanup argv preparation
app: Prepare argv early
elf/app: Delay elf object initialization
elf/app: Pass argc and argv to library initialization
vdso: vdso fast
We want the init of the elf object to use
the tls defined when the new thread is started.
Delay this initialization.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 4 +++-
core/elf.cc| 53 ++---
inclu
This is required by go runtime initialization.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 2 +-
core/elf.cc| 12 ++--
include/osv/elf.hh | 5 +++--
3 files changed, 10 insertions(+), 9 deletions(-)
diff --git a/core/app.cc b/core/app.cc
argv preparation will have to load the vdso compatibility
library and set it's address into the auxv part of argv
_before_ loading the actual program .so.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
linux.cc | 20
1 file changed, 20 insertions(+)
diff --git a/linux.cc b/linux.cc
index e78bb0c..4ab88f3 100644
--- a/linux.cc
+++ b/linux.cc
@@ -24,6 +24,10 @@
#include
#include
#include
+#include
+#i
The previous patch was a raw code move.
This patch takes care of simplifying argv
preparation.
Also get rid of the _argc extra variable.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 58 ++
include/osv/app.h
ficial OSv go main()
entry point. Since the go library init already took argc and
argv as parameters GoMain does not need to bother with them
and hence is specific to Go.:
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc | 3 +++
1 file changed, 3 insertions(+)
diff --git
popfq to restore rflags.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
arch/x64/entry.S | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/arch/x64/entry.S b/arch/x64/entry.S
index cd23cec..04d809d 100644
--- a/arch/x64/entry.S
+++ b/arch/x64/entry.S
@@
This is required by go runtime initialization.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 2 +-
core/elf.cc| 12 ++--
include/osv/elf.hh | 5 +++--
3 files changed, 10 insertions(+), 9 deletions(-)
diff --git a/core/app.cc b/core/app.cc
We want the init of the elf object to use
the tls defined when the new thread is started.
Delay this initialization.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
core/app.cc| 4 +++-
core/elf.cc| 53 ++---
inclu
;
>
> --
> Nadav Har'El
> n...@scylladb.com
>
> On Thu, Oct 20, 2016 at 5:19 PM, Benoît Canet <benoit.canet.contrib@gmail.
> com> wrote:
>
>> Signed-off-by: Benoît Canet <ben...@scylladb.com>
>> ---
>> core/app.cc | 3 +--
>> 1 file change
Add these structure add the end of the list after the
INET and INET6 addresses are filled.
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
libc/network/getifaddrs.c | 71 ---
1 file changed, 67 insertions(+), 4 deletions(-)
diff --git
On Wed, Oct 26, 2016 at 2:03 PM, Nadav Har'El <n...@scylladb.com> wrote:
> Please add "Fixes #787" to the commit message.
>
> Thanks, the patch looks mostly correct, but one question below.
>
> On Wed, Oct 26, 2016 at 2:33 PM, Benoît Canet <benoit.canet.con
Signed-off-by: Benoît Canet <ben...@scylladb.com>
---
Makefile | 2 +-
libc/network/getifaddrs.c | 180 ++
2 files changed, 181 insertions(+), 1 deletion(-)
create mode 100644 libc/network/getifaddrs.c
diff --git a/Make
Hu ho !!! I break java. I will fix it and repost something working.
On Thu, Oct 20, 2016 at 4:19 PM, Benoît Canet <
benoit.canet.cont...@gmail.com> wrote:
> This patch adds support for the go runtime to work
>
> Sample application gives me:
>
> OSv v0.24-206-g2144d63
&
Thanks I will dig this.
On Fri, Oct 14, 2016 at 4:29 PM, Tomasz Grabiec <tgrab...@scylladb.com>
wrote:
>
>
> On Fri, Oct 14, 2016 at 4:14 PM, Benoît Canet <ben...@cloudius-systems.com
> > wrote:
>
>>
>> QEMU on ubuntu 16.04.
>>
>> benoit@alfr
like a wait condition it look strange.
On Fri, Oct 14, 2016 at 3:57 PM, Tomasz Grabiec <tgrab...@scylladb.com>
wrote:
>
>
> On Fri, Oct 14, 2016 at 3:39 PM, Benoît Canet <ben...@cloudius-systems.com
> > wrote:
>
>>
>> Hello list,
>>
>> In the
-5 months ?
>
> Thanks!
>
>
> On Friday 14 October 2016, Benoît Canet <ben...@cloudius-systems.com>
> wrote:
>
>>
>> Hi Ganesh,
>>
>> This sound like a big project given that the OSv tcp stack is heavyly
>> optimised. (Van-Jacobson)
&g
Hi Ganesh,
This sound like a big project given that the OSv tcp stack is heavyly
optimised. (Van-Jacobson)
Maybe you should wait for Nadav the OSv maintainer to come back from
vacation in two week and discuss
that with him.
Best regards
Benoît
On Fri, Oct 14, 2016 at 6:01 AM, GanesH AvacharE
47c7db
0x1147c6cc
0x114cf232
0x2055ff3f
On Fri, Oct 14, 2016 at 3:57 PM, Tomasz Grabiec <tgrab...@scylladb.com>
wrote:
>
>
> On Fri, Oct 14, 2016 at 3:39 PM, Benoît Canet <ben...@cloudius-systems.com
> > wrote:
>
>>
>> Hello list,
>&g
Hello list,
In the process of porting go to OSv I noticed that the Go runtime
initialization stall half the time waiting in a read for /dev/urandom.
The message OSv output is "random: blocking on read.".
It then sometime proceeed to unblock itself a couple dozen of seconds later.
Does it ring
No,
Probably finding a liberaly licensed userspace samba library and wrap it in
an OSv kernel driver like I did for NFS.
If no userpace samba library fit the bill then a port must be done from
another kernel and it will be much harder
to do.
We have an NFS client that work fine. I don't tested
On Sat, Sep 3, 2016 at 10:46 PM, Nadav Har'El <n...@scylladb.com> wrote:
>
> On Fri, Sep 2, 2016 at 5:35 PM, Benoit Canet <benoit.canet.contrib@gmail.
> com> wrote:
>
>> This will allow go program to get the vdso library address.
>>
>> Signed
<benoit.canet.contrib@gmail.
> com> wrote:
>
>> The go c-shared object need the libc to pass them
>> the argc and argv parameters in order to initialize
>> the runtime properly.
>>
>> This is a preparation patch for this.
>>
>> Signed-off-by: B
Hi thanks a lot for your explanations.
So we are suposed to load the c-shared so then run the init().
Do we must pass argc, argv to the .so initializer and that all ?
Or must we then call GoStart after doing this ?
Best regards
Benoît
On Thu, Sep 1, 2016 at 9:54 PM, David Crawshaw
IRELATIVE:
On Wed, Aug 31, 2016 at 4:07 PM, Benoît Canet <ben...@cloudius-systems.com>
wrote:
> This is what I saw in the code managing the flags.
>
> By anyway I think you are right and shared is an "impasse".
>
> On Wed, Aug 31, 2016 at 4:06 PM, Nadav Har'El <n...
This is what I saw in the code managing the flags.
By anyway I think you are right and shared is an "impasse".
On Wed, Aug 31, 2016 at 4:06 PM, Nadav Har'El <n...@scylladb.com> wrote:
>
> On Wed, Aug 31, 2016 at 4:26 PM, Benoît Canet <ben...@cloudius-systems.com
>
buildmode=pie? This should
> build a shared library which can't just (I would hope) take over the fs
> base register on a whim
>
>
>>
>> With this patch and the arch_prctl() stuff I revived from
>> Pekka's previous go work the code now crash after resolving
>&g
Ah sorry he does it with rflag and r11
On Mon, Aug 29, 2016 at 7:10 PM, Benoît Canet <ben...@cloudius-systems.com>
wrote:
>
>
> On Mon, Aug 29, 2016 at 7:08 PM, Avi Kivity <a...@scylladb.com> wrote:
>
>>
>>
>> On 08/29/2016 07:49 PM, Nadav Ha
On Mon, Aug 29, 2016 at 7:08 PM, Avi Kivity wrote:
>
>
> On 08/29/2016 07:49 PM, Nadav Har'El wrote:
>
>> GDB uses DWARF CFI meta-instructions to know how to backtrace through
>> function call frames, and which registers get saved where.
>>
>> This patch fixes the CFI
Should it be just around the "call" instruction ?
On Mon, Aug 29, 2016 at 4:14 PM, Nadav Har'El wrote:
> The AMD64 ABI allows a function to use 128 bytes beyond the bottom of the
> stack without updating the stack pointer - as an optimization known as the
> "red zone". If
onvert forth and back to and from libc. Do it.
>>
>> Signed-off-by: Benoît Canet <ben...@scylladb.com>
>> ---
>> linux.cc | 29 +
>> 1 file changed, 29 insertions(+)
>>
>> diff --git a/linux.cc b/linux.cc
>> index 843d
Go on I'll be busy wrapping the syscalls.
On Mon, Aug 29, 2016 at 2:48 PM, Nadav Har'El wrote:
> By the way, Benoit - with this patch, we save some of the registered saved
> more than once (rcx, rsp), and some space wasted on the stack, all for the
> purpose of building
If you look as the assembly dump of loader.elf you will see save saving rbp
and restoring it is usually the first and last thing function do.
It marks the begining of a frame.
On Mon, Aug 29, 2016 at 12:29 PM, Nadav Har'El wrote:
> Not saving and restoring the rbp register
ug 28, 2016 at 5:38 PM, Benoit Canet <benoit.canet.contrib@gmail.
> com> wrote:
>
>> This C syscall example is just doing exit(42)
>>
>> Signed-off-by: Benoît Canet <ben...@scylladb.com>
>> ---
>> syscall/Makefile | 9 +
>> syscall/mod
Works for me (tm)
Reviewed-by: Benoît Canet <ben...@scylladb.com>
2016-08-25 18:46 GMT+02:00 Nadav Har'El <n...@scylladb.com>:
> This is a trivial hello-world Free Pascal application.
> It does't currently work because of missing syscall instruction support.
> https:
Hello,
I don't know capstan but pekka in copy knows it.
Pekka: could you help them ?
Best regards
Benoît
--
You received this message because you are subscribed to the Google Groups "OSv
Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to
Hi Justin,
You should write proper comments below your commit message for the virgin
reviewer.
Here I have no idea what CPU_COUNT_S is and why you want it.
Additional + proper comments in commits message help us build a sound git
log history.
Best regards
Benoît
On Thu, Aug 18, 2016 at 12:45
I am not sure I understand your question correctly. Had you used some
capstan tutorial previously or the one I pasted you ?
Because these ones generate .vmx files that must be what you are looking
for.
Best regards
2016-08-15 16:26 GMT+02:00 Benoît CANET <ben...@irqsave.net>:
>
&g
9AM +0200, Benoît Canet wrote:
> > Hi,
> >
> > I am working on a slowness bug affecting the EU project.
> >
> > In this test HDFS a java application does a write test 7 times slower
> than
> > Linux.
> >
> > The people doing the test sent me some tc
91 matches
Mail list logo