[PATCH] Do not include libosv.so in Java module

=ramfs (note the fs=ramfs, causing all the files to be included in the kernel executable) fails with strip: 'libosv.so': No such file This patch simply removes libosv.so from the java module. Nothing seems to rely on it. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- module

[PATCH] build: allow easily increasing the maximum allowed kernel size

a certain length), and in Makefile (which verifies that the kernel isn't too large). Fixes #872. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- fastlz/lzloader.cc | 10 -- Makefile| 16 +--- arch/x64/boot16.S | 2 +- arch

Re: [PATCH 2/3] Ported subset of Hyper/V vmbus driver code (hyperv.cc) from FreeBSD to C++ and adapted where necessary to use OSv apis

Looks mostly good. Some minor comments below. On Fri, May 26, 2017 at 8:37 AM, Waldemar Kozaczuk wrote: > Here is the list of specific changes: > - where required by C++ compiler applied appropriate C++ cast > (static_cast, reinterpret_cast, const_cast) > - where

Re: HyperV/Azure: work in progress status

ot;TODO" and only go back to it after you make the network (for example) work. -- Nadav Har'El n...@scylladb.com On Fri, May 19, 2017 at 6:03 PM, Waldek Kozaczuk <jwkozac...@gmail.com> wrote: > The essence of my changes is this new clock class - https://github.com/ > wkozaczuk/osv/blob/h

Re: HyperV/Azure: work in progress status

On Fri, May 19, 2017 at 11:54 PM, Waldek Kozaczuk wrote: > What do you mean by "exit"? Is it because it calls a function? Would > inlining help? > "VMX" (virtual machine extensions) is Intel's hardware support for virtualization. What makes it efficient is that the guest

Re: HyperV/Azure: work in progress status

Hi, On Fri, May 19, 2017 at 7:52 AM, Waldek Kozaczuk wrote: > I finally managed to fully boot and execute simple C hello work program on > Hyper/V. > That's great! With the patch I sent yesterday allowing a kernel of any size, you should now be able to run any OSv image

[PATCH] stdio: fix fread() from block device

once for the entire readv() scattered buffer, as expected. So we don't need this workaround, which can only hurt performance, and in the present case, prevent fread() from working on a block device. Fixes #877 Signed-off-by: Nadav Har'El <n...@scylladb.com> --- tests/tst-fread.cc

Re: [PATCH] stdio: fix fread() from block device

On Tue, Jun 13, 2017 at 5:54 PM, Justin Cinkelj wrote: > Patch seems to work, please apply. > > Instead of assert(), libcdio now complains if I ask to extract invalid > filename. > If you tried to read a non-existant file, it should have also failed this way

Re: HyperV/Azure: work in progress status

On Sun, May 7, 2017 at 9:09 AM, Waldek Kozaczuk wrote: > Hyper/V VMBUS seems be using inter processor interrupts delivered by > *Synthetic > Interrupt Controller (SynIC). * > > The BSD drivers use lapic_ipi_alloc() which per this patch description ( >

Re: [PATCH osv-apps] Bump Node.js version to 6.10.3

versions, but I suspect that this might not be necessary after I added support for nearbyint(). I also had several other reservations on that patch, which is why it wasn't committed. Did you also try more recent versions and failed? Or did you really want to run 6.10 in particular? -

Re: [PATCH osv-apps] Bump Node.js version to 6.10.3

On Mon, May 8, 2017 at 2:48 PM, Nadav Har'El <n...@scylladb.com> wrote: > Thanks. I committed this. > > In February, Rean Griffith <rg2...@caa.columbia.edu> sent patch titled > "Generalize > nodejs build to support recent versions (6.9.5 LTS and 7.5.0 latest)"

Re: httpserver /fs/df and block size

:-) -- Nadav Har'El n...@scylladb.com On Wed, Sep 13, 2017 at 8:56 PM, Waldek Kozaczuk <jwkozac...@gmail.com> wrote: > Currently httpserver returns disk usage information in blocks like this > snippet demonstrates: > > static void fill_dfstat(DFStat& dfstat, const osv::moun

Re: [PATCH] iperf: Update to recent iperf version

I found https://sourceforge.net/projects/iperf2/files/ -- Nadav Har'El n...@scylladb.com On Mon, Sep 18, 2017 at 2:53 PM, Nadav Har'El <n...@scylladb.com> wrote: > > On Mon, Sep 18, 2017 at 2:51 PM, Nadav Har'El <n...@scylladb.com> wrote: > >> >> On Mon, Sep

Re: [PATCH] iperf: Update to recent iperf version

On Mon, Sep 18, 2017 at 2:51 PM, Nadav Har'El <n...@scylladb.com> wrote: > > On Mon, Sep 18, 2017 at 1:28 PM, Benoît Canet <ben...@scylladb.com> wrote: > >> >> I think it's an enthousiast that package iperf for easy distribution. >> >> What

Re: [PATCH] iperf: Update to recent iperf version

On Fri, Sep 15, 2017 at 3:58 PM, Benoît Canet < benoit.canet.cont...@gmail.com> wrote: > This one does compile on recent host. > Three weeks ago Justin noticed the same problem, and we fixed this by a trivial fix to iperf's autoconf (commit fd93d5b996720e60e617e894dccb028ef89e8ccb in apps.git).

Re: [PATCH] iperf: Update to recent iperf version

On Mon, Sep 18, 2017 at 1:28 PM, Benoît Canet wrote: > > I think it's an enthousiast that package iperf for easy distribution. > > What worry me is that the official mirror > https://github.com/esnet/iperf does not distribute version 2 that we need > for osv. > I just

[PATCH] Fix crash in read_partition_table()

to process the read data, even though it wasn't, and crashed on this processing or on the brelse() at its end. In this case, the read_partition_table() should just do nothing, and certainly not crash. Refs #903. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- fs/devfs/device.cc | 5 ++

[PATCH 2/2] lzloader: fix memset() implementation

nd results in endless recursion and a hanging boot... So let's implement memset() using the base_memset() we already have in libc/string/memset.c. Fixes #913. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- fastlz/lzloader.cc | 15 ++- 1 file changed, 10 insertions(+), 5 de

[PATCH 1/2] memset_base(): avoid warnings in stricter C++ compiler

/lzloader.cc. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- libc/string/memset.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/libc/string/memset.c b/libc/string/memset.c index 0d7f1c9..c42d00c 100644 --- a/libc/string/memset.c +++ b/libc/string/memset.c @@ -9,13

Re: [RFC] namespace: enable per namespace alarm/signal

"partially" work as it does now, for only some applications? . -- Nadav Har'El n...@scylladb.com On Wed, Mar 8, 2017 at 9:22 AM, Justin Cinkelj <justin.cink...@xlab.si> wrote: > Before patch, netperf netserver.so and netperf.so could not be run in > the same VM. The netserver r

Re: [PATCH v2 2/2] namespace: fix signal handling called from unduplicated code

On Mon, Oct 16, 2017 at 4:30 AM, Justin Cinkelj wrote: > The previous patch (namespace: enable per namespace alarm/signal) > implemented per elf namespace signals for common usecase of alarms. > > But, as Nadav explained, it failed to allow application in a separate > ELF

Re: [PATCH v2 1/2] namespace: enable per namespace alarm/signal

On Mon, Oct 16, 2017 at 4:30 AM, Justin Cinkelj wrote: > From: Justin Cinkelj > > 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". > > The patch

Re: LinuxKit from Docker

On Thu, Sep 7, 2017 at 8:50 PM, Waldek Kozaczuk wrote: > Is linuxkit from Docker (https://github.com/linuxkit/linuxkit) going to > kill unikernels altogether? What do you think? > Hi, This approach can definitely work - I have done something like this 17 years ago (even

Re: [PATCH] iperf: fix compile on gcc 6.3

Hi, sorry, I didn't see this patch, and committed a different patch (c8b1906cf59d73a7cb39926c53dc4e136611d4f5) at about the same time, doing the same thing in a different way. -- Nadav Har'El n...@scylladb.com On Tue, Aug 29, 2017 at 11:37 AM, Justin Cinkelj <justin.cink...@xlab.si>

Re: [PATCH v2 2/6] libc: fix of-by-one error in stresep

On Thu, Aug 24, 2017 at 8:29 PM, Justin Cinkelj wrote: > memcpy had of-by-one error. > Also, remove unneeded defines. > > Signed-off-by: Justin Cinkelj > --- > libc/string/stresep.c | 16 ++-- > 1 file changed, 2 insertions(+), 14

Re: [PATCH v3 0/6] Allow spaces in loader options

On Wed, Sep 13, 2017 at 3:05 PM, Justin Cinkelj wrote: > Compared to v2 patch: > The stresep.c file is now first only added, and included into Makefile > after removing unneeded define and include namespace.h. > The memcpy bug from v2 patch was reported upstrem to netbsd,

Re: [PATCH v3 1/2] Split httpserver module into httpserver-api and httpserver-html5-gui

Thanks, I committed (sorry for the delay). As usual when git submodules are involved, I had somewhat of a mess applying this patch (needed to manually remove some old directories and "git submodule update"), I hope users and jenkins will not be bit by this. -- Nadav Har'El n...@sc

Re: [PATCH 4/6] command line: use loader_parse_cmdline

On Thu, Aug 24, 2017 at 7:56 PM, Justin Cinkelj wrote: > Parse loader options with loader_parse_cmdline instead of strtok_r. > The cmdline passed in prepare_commands to osv::parse_command_line > is not reassebled from already parsed argv anymore, as app_cmdline > contains

Re: [PATCH v2 3/6] command line: add loader_parse_cmdline with tests

On Thu, Aug 24, 2017 at 8:29 PM, Justin Cinkelj wrote: > loader_parse_cmdline accepts input str - OSv commandline, say: > --env=AA=aa --env=BB=bb1\ bb2 app.so arg1 arg2 > > The loader options are parsed and saved into argv, up to first > not-loader-option > token. argc is

Re: lzloader compile issue

On Sun, Oct 1, 2017 at 5:00 AM, Rick Payne wrote: > Hi, > > I’m still having to apply this patch to get OSv to work on my Ubuntu 16.10 > box. Without it, the produced image fails to start even for the initial > cpio phase. Its clearly some optimisation issue with gcc 6.2

Re: [PATCH] procfs: Fix procfs

On Sun, Oct 1, 2017 at 5:12 PM, Benoît Canet wrote: > /proc/stat was in the wrong place. > Move it. > > Signed-off-by: Benoît Canet > --- > fs/procfs/procfs_vnops.cc | 3 +-- > 1 file changed, 1 insertion(+), 2 deletions(-) > > diff --git

Re: SO_BINDTODEVICE

On Thu, Aug 17, 2017 at 10:54 PM, Rick Payne (Offshore) < ri...@rossfell.co.uk> wrote: > > > On 17 Aug 2017, at 06:43, Rick Payne (Offshore) > wrote: > > > > Does carp run on Freebsd or is it just OpenBSD? That has to solve a > similar issue, namely sending announcements to

Re: [PATCH] cmdline: allow parameters with escaped spaces

On Mon, Aug 21, 2017 at 7:01 PM, Justin Cinkelj wrote: > The strtok_r doesn't allow escaping the input parameters, which means > each space is considered a separator. The boost::tokenizer supports > escaping. This commit replaces strtok_r with boost::tokenizer. The >

Re: [PATCH] cmdline: allow parameters with escaped spaces

On Tue, Aug 22, 2017 at 2:48 PM, Justin Cinkelj wrote: > > What does it not fix? > The $892 mentions environ varibales with spaces or empty one. And command > line parsing is done in loader.cc to get loader options, and in > commands.cc, to get argv/argc for app, and to

Re: SO_BINDTODEVICE

On Mon, Aug 21, 2017 at 9:01 PM, Rick Payne (Offshore) wrote: > > I don’t know about vmxnet3 driver as I’m not familiar with the hypervisor > that provides that. Can you point me at the code for the ‘other side’ for > that? > vmxnet3 is VMWare's driver. I don't think we

Re: [PATCH v2] Enhanced scripts/build and related Python scripts to support export files that make up an image versus uploading them to an image

Thanks. I committed this patch. One thing I didn't understand is the comment: "The only current limitation is support for links in OSv". I see you did handle symlinks... Is this an old comment? -- Nadav Har'El n...@scylladb.com On Sat, Aug 19, 2017 at 7:51 PM, Waldemar Kozaczu

Re: nginx failing to handle concurrent connections

On Fri, Sep 1, 2017 at 7:08 PM, Waldek Kozaczuk wrote: > I changed > int sendfile(int out_fd, int in_fd, off_t *_offset, size_t count) > to > ssize_t sendfile(int out_fd, int in_fd, off_t *_offset, size_t count) > > and all works like a charm. > Wow, thanks, great

Re: nginx failing to handle concurrent connections

On Sun, Sep 3, 2017 at 10:26 AM, Nadav Har'El <n...@scylladb.com> wrote: > > On Fri, Sep 1, 2017 at 7:08 PM, Waldek Kozaczuk <jwkozac...@gmail.com> > wrote: > >> I changed >> int sendfile(int out_fd, int in_fd, off_t *_offset, size_t count) >> to >&g

Re: [COMMIT osv master] Waldek Kozaczuk discovered that nginx, using sendfile(), is experiencing

On Sun, Sep 3, 2017 at 10:55 AM, Commit Bot <b...@cloudius-systems.com> wrote: > From: Waldemar Kozaczuk <jwkozac...@gmail.com> > Committer: Nadav Har'El <n...@scylladb.com> > Branch: master > > Waldek Kozaczuk discovered that nginx, using sendfile(), is exp

Re: [PATCH] new object loading exceptions in core/elf.cc

On Wed, Oct 18, 2017 at 1:42 AM, wrote: > As a way of getting up to speed on the codebase, finish up the remaining > work on issue #904. Add new exception type and catch that where necessary. > If you put the text "Fixes #904" somewhere inside the commit message, github will

Re: I don't know how to use memcached image.

On Fri, Nov 24, 2017 at 11:30 AM, wrote: > Thanks for your reply. > > I did as the instruction: > wande@ubuntu:~/osv$ sudo scripts/run.py -e "memcached -u root" --forward > "tcp:1121::1121" > OSv v0.24-467-gd52cb12 > eth0: 192.168.122.15 > Memcached for OSV, based on

Re: Enhance ramfs/zfs-less images

On Wed, Nov 22, 2017 at 8:01 AM, Waldek Kozaczuk wrote: > I managed to implement my change to make msf_read() use device strategy > routing directly to bulk read however many blocks at once the uio parameter > to mfs_read() specifies and as expected now the java run takes ~

Re: I don't know how to use memcached image.

On Wed, Nov 22, 2017 at 5:24 PM, Waldek Kozaczuk wrote: > You probably need to expose the port 1121 by adding forward option like so: > > scripts/run.py -e "memcached -u root" --forward "tcp:1121::1121" > Right. By default, run.py runs the image you just built

Re: Enhance ramfs/zfs-less images

s (the buffers) in a std::list or boost::intrusive::list and every time you find a buffer, move it to the end of the list. James Root use in his mfs_cache.hh/cc "next" and "prev" pointers and C style code. > On Wednesday, November 22, 2017 at 1:42:01 AM UTC-5, Nadav Har'El

Re: I don't know how to use memcached image.

On Thu, Nov 30, 2017 at 3:32 PM, wrote: > I started with > "scripts/run.py -e "memcached -u root -p 1121" --forward "tcp:1121::1121"" > which I add "-p 1121". > Then it worked. > Before I added "-p",it showed me > "Connection closed by foreign host" > Thanks very much. > >

Re: Some unit tests mask failures

On Sun, Dec 17, 2017 at 6:53 PM, Waldek Kozaczuk wrote: > As I have been working on unit tests around read-only filesystem I noticed > that some of the unit tests (look for 'debug\(.*PASS' under tests) report > success even though I know for sure (got it through

Re: Enhance ramfs/zfs-less images

>>>> Network initialized: 53.12ms, (+0.24ms) >>>> pvpanic done: 53.24ms, (+0.12ms) >>>> pci enumerated: 57.59ms, (+4.35ms) >>>> drivers probe: 57.59ms, (+0.00ms) >>>> drivers loaded: 100.55ms, (+42.96ms) >>>> ZFS mount

Re: Enhance ramfs/zfs-less images

>> See my further question below. >> >> On Monday, November 20, 2017 at 2:26:45 PM UTC-5, Nadav Har'El wrote: >>> >>> Hi, I read through your interesting measurements in this thread. >>> >>> I agree with you that the problem is likely the rea

Re: Enhance ramfs/zfs-less images

On Tue, Nov 21, 2017 at 3:43 PM, Waldek Kozaczuk wrote: > > I am still interested how to make OSv page manager to request multiple > pages. > It's been a while since I touched any of this code :-( Unfortunately it's not yet clear to me how to do this cleanly. We could

Re: AWS new KVM hypervisor

On Fri, Nov 10, 2017 at 3:24 PM, Avi Kivity wrote: > OSv will need NVMe and ena drivers. > Why is that, have you seen documentation (or saw in your own tests) that they will not support virtio-blk and virtio-net on their KVM? > > On Nov 10, 2017 15:07, "Waldek Kozaczuk"

Re: Enhance ramfs/zfs-less images

On Wed, Nov 15, 2017 at 5:21 AM, Waldek Kozaczuk wrote: > A better solution could be to not have a read-write ramdisk at all, but > rather a simple read-only filesystem on the disk image where we only read > file blocks as they are needed, and also cache them somehow for

Re: Slow to upload thousands of tiny files over cpiod

On Sat, Nov 4, 2017 at 5:14 AM, Waldek Kozaczuk wrote: > Miha, > > It looks like you are right. When I disabled printing of files names in > cpiod it made some difference when using QEMU and huge difference when > using VirtualBox on OSX. With QEMU on average I was able to

Re: Memory limit?

On Mon, Nov 6, 2017 at 10:26 PM, Rick Payne (Offshore) wrote: > > Out of memory: could not reclaim any further. Current memory: 5122256 Kb > > > > This suggests there was 5GB free while the allocation failed. > > This *can* be a fragmentation issue (e.g., you asked for a 1

Re: Memory limit?

On Tue, Nov 7, 2017 at 10:35 AM, Rick Payne wrote: > Hi, > > > I am not sure how this will help, as the later malloc() can still fail > when it wants to allocate physically-contiguous memory. > > > > One hack you can try to fix https://github.com/cloudius- >

Re: Remotely set cmdline

instance is first > connected to internal network. Then later must user manually assign > a "floating IP" to it to make it accessible from the wild. > > Thanks for a reply, I'm glad you noticed the problem. > > > Dne sreda, 08. november 2017 17.06.31 UTC+1 je oseba N

Re: Remotely set cmdline

On Thu, Nov 9, 2017 at 3:16 PM, Waldek Kozaczuk wrote: > Yep I had similar thoughts about security. I would imagine that on AWS one > would have a "build" EC2 instance where he/she would run capstan and spin a > "stem" OSv instance to upload files to and take snapshot to

Re: Remotely set cmdline

On Wed, Nov 8, 2017 at 9:59 AM, wrote: > Nice one, this should work (with limitations as you described, but I don't > think they really bother us). So the usage would be: > 1) prepare base unikernel with* '/tools/cpiod.so --prefix /'* line in > /run/init > Something

Re: Memory limit?

On Thu, Nov 2, 2017 at 6:38 AM, Rick Payne wrote: > > I’m stressing OSv a bit, and though I start the VM with 10G of memory, it > seems to fail after just over 5GB. Is there a limit that I’m hitting, or > perhaps my memory usage is fragmenting things too much? > > Out of

Re: OSv image under ProxMox VE5.1

On Mon, Oct 30, 2017 at 2:45 PM, Player, Timmons wrote: > I’ve run into the same issue under VMware on VM’s that lack a serial > port. I’ve used the following patch locally with success… > > > From 44320eb6d9bc042ca066073f8c83cda27139b1e6 Mon Sep 17 00:00:00 2001 >

Re: OSv image under ProxMox VE5.1

On Thu, Nov 2, 2017 at 4:35 PM, Nadav Har'El <n...@scylladb.com> wrote: > On Mon, Oct 30, 2017 at 2:45 PM, Player, Timmons < > timmons.pla...@spirent.com> wrote: > >> I’ve run into the same issue under VMware on VM’s that lack a serial >> port. I’ve used the fol

Re: how to run some .so or jar

On Tue, Nov 28, 2017 at 1:56 PM, wang Yu wrote: > i am fresh man in osv, and i have a question about it; > if we have some dyminic library .so, i want to how to run? > for example, a.so have main enter, and it want to call the func_b() > in b.so(we don't want to recompile

Re: [COMMIT osv master] Enhanced tests to allows testing rofs/ramfs image

n again a ZFS build, even though the only thing changed is the command line? If this doesn't work, then Jenkins won't work properly (I believe it doesn't bother with a "make clean" every time). > > Waldek > > On Sunday, May 13, 2018 at 5:44:29 PM UTC-4, Bot Droid wrote: >> &g

Re: osv-apps: HEAD detached at 7d9b71e

On Wed, May 9, 2018 at 7:55 PM, Geraldo Netto wrote: > Dear Friends, > > Just a quick thing, I was syncing my local OSv repos > and I noticed something with osv-apps: > > netto@besta01 ~/Desktop/osv/osv-clean $ git pull > remote: Counting objects: 7, done. > remote:

Re: [COMMIT osv master] Gracefully handle the case of missing libvsdo.so

On Sun, May 13, 2018 at 7:02 AM, Commit Bot wrote: > From: Waldemar Kozaczuk > Committer: Waldemar Kozaczuk > Branch: master > > Gracefully handle the case of missing libvsdo.so > > Current implementation of

Re: [PATCH v3] Implement switch to separate SYSCALL call stack

Hi, thanks for being so insistent on this important patch. I really wanted to commit this version, but unfortunately after another reading, I'm really worried if what you're doing with rbx is safe. Please see my comments below. Sorry about dragging this on for so long :-( On Fri, Mar 16, 2018

Re: Library unikernel or modularization of OSv kernel

On Sun, May 6, 2018 at 7:52 AM, Dor Laor wrote: > On Sat, May 5, 2018 at 9:48 AM, Waldek Kozaczuk > wrote: > >> What if there was a way early into the booting process to load specific >> features of OSv, that right now are linked in as part of kernel, as

Re: [PATCH v5] Implement switch to separate SYSCALL call stack

-- Nadav Har'El n...@scylladb.com On Tue, May 15, 2018 at 2:59 PM, Waldemar Kozaczuk <jwkozac...@gmail.com> wrote: > This patch implements separate syscall call stack needed > when runtimes like Golang use SYSCALL instruction to execute > system calls. More specifically in case

[PATCH 2/2] poll: don't return POLLNVAL from poll_no_poll()

, adding an extra bit not listed, like EPOLLRDHUP, caused POLLNVAL to be returned. Worse, when using epoll, it did not expect to ever get back POLLNVAL from poll(), and caused an assertion failure. Refs #971. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- core/poll.cc | 11 +--

[PATCH 1/2] epoll: don't pass unknown bits to poll()

poll() inplementations, and in particular poll_no_poll() which caused an assertion failure. Refs #971. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- core/epoll.cc | 9 - 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/core/epoll.cc b/core/epoll.cc index f425c4ad..42

Re: [PATCH v6] Implement switch to separate SYSCALL call stack

On Tue, May 15, 2018 at 10:20 PM, Waldemar Kozaczuk wrote: > This patch implements separate syscall call stack needed > when runtimes like Golang use SYSCALL instruction to execute > system calls. More specifically in case of Golang tiny stacks > used by coroutines are not

[PATCH] epoll: fix regression in EPOLLET on network sockets

e for more information). So although the EPOLLET bit is *not* traditionally supported by the poll() method, we in OSv we do need to pass it on to poll(). Other bits, namely EPOLLONESHOT, we still don't need to pass. Fixes #974. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- co

[PATCH] epoll: test polling of /proc file

in #971). The test can now be run on both Linux and OSv, and succeds on both. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- tests/tst-epoll.cc | 40 1 file changed, 40 insertions(+) diff --git a/tests/tst-epoll.cc b/tests/tst-epoll.cc index 6a

Re: [PATCH v4] Implement switch to separate SYSCALL call stack

Thanks, and sorry for all the round-trips... I have one last (I hope) issue below (two comments, actually): -- Nadav Har'El n...@scylladb.com On Tue, May 15, 2018 at 8:01 AM, Waldemar Kozaczuk <jwkozac...@gmail.com> wrote: > This patch implements separate syscall call stack need

Re: Build failed in Jenkins: osv-build #1362

nk...@scylladb.com> wrote: > >> See <http://jenkins.cloudius-systems.com:8080/job/osv-build/ >> 1362/display/redirect?page=changes> >> >> Changes: >> >> [Nadav Har'El] Clean mutex_init()/mutex_destroy() warnings in gcc 8 >> >> ---

Re: [COMMIT osv master] httpserver: do not use Boost Regex library

Ok, so after the barrage of patches I committed today, OSv finally builds and works, mostly, on Fedora 28. Most of the things I tested appear to work now. But "make check" still fails on the weird "tracing_smoke_test" - see issue 964, maybe because of some gdb interaction.

Re: [PATCH] Remove soisconnected() from syncache_socket()

On Thu, Apr 26, 2018 at 8:43 AM, Waldemar Kozaczuk wrote: > This patch fixes an intermittent issue happening in TCP stack code > during session 3-way handshake protocol due to inconsistent state > of transmission control block (TCB) data structure. Specifically > right

Re: [PATCH] Remove soisconnected() from syncache_socket()

On Mon, Apr 30, 2018 at 1:56 AM, Waldek Kozaczuk <jwkozac...@gmail.com> wrote: > See my comments below. > > On Sunday, April 29, 2018 at 6:41:01 PM UTC-4, Nadav Har'El wrote: > >> >> On Thu, Apr 26, 2018 at 8:43 AM, Waldemar Kozaczuk <jwkoz...@gmail.com&g

Re: Build failed in Jenkins: osv-build-nightly #1451

On Fri, Apr 27, 2018 at 5:41 PM, Waldek Kozaczuk wrote: > This time tst-unordered-ring-mpsc failed (line 60). > What do you mean "this time"? In the message you quote below, the error is, "as usual", in tst-tcp-nbwrite.so. Or do you mean that you ran tests on your own

Re: Build failed in Jenkins: osv-build-nightly #1451

On Fri, Apr 27, 2018 at 6:19 PM, Waldek Kozaczuk wrote: > My theory might be right. > > I ran simple experiment by adding this before first assert > in test_when_ring_gets_full test: > sched::thread::pin(sched::cpus[0]); > > and then before assert(!ring.push(5)); >

Re: golang-example fails with general protection or page fault

On Thu, Dec 28, 2017 at 7:09 PM, Waldek Kozaczuk wrote: > Also I have noticed my stack trace does not show Golang symbol names like > some example stacktraces in #850 do have. What am I missing to make the > stack traces to show Golang symbols? I dit verify with readelf

Re: golang-example fails with general protection or page fault

On Sat, Jan 6, 2018 at 7:22 AM, Waldek Kozaczuk wrote: > I applied Hawxchen's patch and at first the httpserver app would crash > with following type of error. It would be quite consistent and sometimes it > would show golang stack trace but sometimes not like below.stack

[PATCH] x64/processor.hh: fix wrongly named variable

e register - not addresses. This patch only changes variable names - it should have no functional significance. Signed-off-by: Nadav Har'El <n...@scylladb.com> --- arch/x64/processor.hh | 20 ++-- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/arch/x64/process

Re: [PATCH 1/2] * avoid downloading tomcat if it already exists

On Thu, Jan 11, 2018 at 2:17 PM, Geraldo Netto wrote: > Hello Nadav/All, > > Thanks for your review, I think we can also check with md5/sha1 > Will get back to you as soon as I finish it :) > > Out of my curiosity, do we have any consolidated list of to do/fix me/wish >

Re: golang-example fails with general protection or page fault

On Wed, Jan 10, 2018 at 6:24 AM, Waldek Kozaczuk wrote: > Your observation was correct. It worked. At least I have not seen same > errors in any of my tests after the change. Thanks !!! > Great. The approach of adding another function "syscall_wrapper" makes system calls a

Re: [PATCH 1/2] * avoid downloading tomcat if it already exists

On Wed, Jan 10, 2018 at 2:45 PM, geraldo netto wrote: > Signed-off-by: geraldo netto > --- > tomcat/GET | 7 ++- > 1 file changed, 6 insertions(+), 1 deletion(-) > > diff --git a/tomcat/GET b/tomcat/GET > index b96d963..35351d8 100755 > ---

Re: [PATCH v2] * clean ROOTFS if it already exists

This looks like a patch to your previous patch. Can you please send one patch? Thanks. -- Nadav Har'El n...@scylladb.com On Wed, Jan 10, 2018 at 3:26 PM, geraldo netto <geraldone...@gmail.com> wrote: > Signed-off-by: geraldo netto <geraldone...@gmail.com> > --- > tomc

Re: golang-example fails with general protection or page fault

ernals-part-5- > runtime-bootstrap-process.html. I am still digesting all that information > to see where in golang sourcecode put a debug statement that would prove > that indeed the underlying issue is as described in #808 > I think it will be hard to prove with printouts. If

Re: Possible thread synchronization issue in futex implementation

On Thu, Jan 11, 2018 at 1:28 AM, Waldek Kozaczuk wrote: > I added more debug statements and now I think it might be a deadlock > scenario between 3 threads. It might be a bug in Golang but why it work > fine in Linux. > I'll look at your traces in a few minutes, but before

Re: Possible thread synchronization issue in futex implementation

On Thu, Jan 11, 2018 at 1:28 AM, Waldek Kozaczuk wrote: > I added more debug statements and now I think it might be a deadlock > scenario between 3 threads. It might be a bug in Golang but why it work > fine in Linux. > > Here is a newest debug output and latest linux.cc.

Re: golang-example fails with general protection or page fault

t; lea 128(%rsp), %rsp > #else > xchg %rsp, %fs:16 > #endif > > # jump to rcx where the syscall instruction put rip > # (sysret would leave rxc cloberred so we have nothing to do to > restore it) > jmpq *%rcx >.cfi_endproc > .size syscall_e

Re: [PATCH] memory: respect _bitset length bounds in page_range_allocator

On Tue, Jan 16, 2018 at 12:31 AM, Player, Timmons < timmons.pla...@spirent.com> wrote: > So, it is possible to reproduce the issue with a unit test, however it > requires specifying an explicit memory size to the run script. > > The issue only occurs when the bitset size is an exact multiple of >

Re: golang-example fails with general protection or page fault

On Mon, Jan 15, 2018 at 11:25 PM, Waldek Kozaczuk wrote: > Yep I will try to follow your advice about the "least resistance" order. > > I could not find anything about " TLS-related" or "the order of when the > TLS variables were allocated or initialized" related patches.

Re: [PATCH] Added number of syscalls to support golang

On Mon, Jan 15, 2018 at 10:22 PM, Waldek Kozaczuk wrote: > I forgot about mentioning Benoit. Shall I add "Signed by" him line or > mention that he is the author of most changes? > I'm not sure "signed-off-by" is supposed to indicates authorship. You can perhaps just add a

Re: golang-example fails with general protection or page fault

On Tue, Jan 16, 2018 at 12:03 PM, Nadav Har'El <n...@scylladb.com> wrote: > > On Mon, Jan 15, 2018 at 11:25 PM, Waldek Kozaczuk <jwkozac...@gmail.com> > wrote: > >> Yep I will try to follow your advice about the "least resistance" order. >> &g

Re: [PATCH] linux: arch_prctl() system call

relevant. Perhaps it was relevant to one Go version or one runtime mode (e.g., static compilation but not c-shared), but no longer is relevant to current Go and c-shared. If you don't see any messages about this system call being missing, we can drop it. > > On Wednesday, August 31, 2016 at

Re: [PATCH] memory: fix off by 1 error in page_range_allocator

On Fri, Jan 12, 2018 at 8:23 PM, Timmons C. Player < timmons.pla...@spirent.com> wrote: > The page_range_allocator free() method neglected to subtract > 1 from the index when checking the end of the page range in > the _bitmap. Fix the calculation. > > Note: with this change, the end index

Re: [PATCH] fix kafka mirror url (error 404)

timeout: 6000 until a few seconds later, the application shuts down. Were you able to use this package? How? -- Nadav Har'El n...@scylladb.com On Fri, Jan 12, 2018 at 4:31 AM, geraldo netto <geraldone...@gmail.com> wrote: > Signed-off-by: geraldo netto <geraldone...@gmail.com>

Re: [PATCH] Apache-Kafka: state that Apache ZooKeeper is required

On Sat, Jan 20, 2018 at 7:13 PM, geraldo netto wrote: > Signed-off-by: geraldo netto > --- > apache-kafka/README | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/apache-kafka/README b/apache-kafka/README > index

Re: [PATCH] remove erjang due the following reasons: - the download url is from dropbox and is unavailable [1] - the last version of erjang implementation is dated from 2016 and it seems abandoned [2]

nd out why. So I'll commit your patch and remove it. -- Nadav Har'El n...@scylladb.com On Sat, Jan 20, 2018 at 6:51 PM, geraldo netto <geraldone...@gmail.com> wrote: > [1] https://dl.dropboxusercontent.com/u/1360473/erjang-R16B01.jar > [2] https://github.com/trifork/erjang > >

Re: Unsupported hardware instructions?

On Fri, Jan 19, 2018 at 8:05 PM, Geraldo Netto wrote: > Hello Friends, > > Interesting, if I pass "-cpu phenom" it boots clearly: > I think this is a qemu bug that was solved a while back - e.g., https://bugzilla.redhat.com/show_bug.cgi?id=1326721 > netto@spo01

<    1   2   3   4   5   6   7   8   9   10   >