The code in pipe_buffer.cc had the meaning of POLLHUP and POLLRDHUP
reversed, and as a result POLLHUP wasn't correctly returned when polling
a read-end of the pipe whose write-end is closed.

This fixes the mixup, and as a result the test added to tst-pipe.cc in
the previous commit, and also another one added in this patch, all pass.

We are still incompatible with Linux with what we do regarding POLLRDHUP.
This will be a separate patch.

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Our poll_wake() code ignored calls with the POLLHUP event, because
the user did not explicitly ask for this event. This causes a poll()
waiting on read from a pipe whose write side closes not to wake up.

This patch adds a test for this case in tst-pipe.cc, and fixes the
bug by adding to the poll structure's _events also ~POLL_REQUESTABLE,
i.e., any bits which do not have to be explicitly requested by the
user (POLL_REQUESTABLE is a new macro defined in this patch).

After this patch, poll() wakes as needed in the test (instead of just
hang), but returns the wrong event because of another bug which will
be fixed in a separate patch.

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The assertion in object::relocate_pltgot uses assignment instead of
comparison.  Fix that up.

Spotted by Coverity.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

JavaVMOption has extraInfo member that's not initialized in mkoption().
It only used by few special case options that provide a function pointer
hook to the JVM so it's harmless but fix it anyway.

Spotted by Coverity.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

This will generate cscope index for the source files.

Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Currently, OSv's "make" builds both the kernel, using a normal Makefile,
and the sample management framework, using the "Gradle" tool.

We carefully hid Gradle's progress under one output line "GRADLE", but
this leads to the "make" process hanging for a long time - up to several
minutes, and all the puzzled user sees is the line "GRADLE". Users who
didn't go for coffee will likely kill the build at this point ;-) Even
worse, when Gradle fails, or just hangs (e.g., a bad or slow network
connection), the user won't even know why.

So let's just run Gradle normally, and let the user see its full output.
This output is rather pretty and organized, so no real reason to hide it.

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Each of the "tags" and "TAGS" make targets was done in a different way,
causing each to have a different problem:

 1. "tags" (ctags) used the "-L" option which turns out is nonportable
    (only available in Exuberant Ctags).
 2. "TAGS" (etags) ran the etags command separately for each source file,
    slowing it down.

The best of both worlds is to use xargs to have ctags/etags operate on
multiple files in each run using xargs. Because we cannot be sure xargs
will run ctags/etags only once, we must delete the file first and use
 the "-a" (append) option.

Also, this patch reduces code duplication - there is now one rule for
both "tags" and "TAGS" targets that uses the correct tool (ctags or etags,
respectively).

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Update devices-and-drivers.txt

to do. However, it did several things wrong that this patch fixes. Thanks
to Paolo Bonzini for finding these problems (see issue #35).

1. When poll() returned a bare POLLHUP, without POLLIN, our select() didn't
return a read event. But nothing in the manpages guarantees that POLLHUP
is accompanied by POLLIN, and some special file implementations might
forget it. As an example, in Linux POLLHUP without POLLIN is common.
But POLLHUP on its own already means that there's nothing more to read,
so a read() will return immediately without blocking - and therefore
select() needs to turn on the readable bit for this fd.

2. Similarly, a bare POLLRDHUP should turn on the writable bit: The
reader on this file hug up, so a write will fail immediately.

3. Our poll() and select() confused what POLLERR means. POLLERR does not
mean poll() found a bad file descriptor - there is POLLNVAL for that.
So this patch fixes poll() to set POLLNVAL, not POLLERR, and select()
to return with errno=EBADF when it sees POLLNVAL, not POLLERR.

4. Rather, POLLERR means the file descriptor is in an error state, so every
read() or write() will return immediately (with an error). So when we see
it, we need to turn on both read and write bits in this case.

5. The meaning of "exceptfds" isn't clear in any manual page, and it
seems there're a lot of opinions on what it might mean. In this patch I
did what Paolo suggested, which is to set the except bit when POLLPRI.
(I don't set exceptfds on POLLERR, or any other case).

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Fixed some minor spelling mistakes

Now that we have an efficient interrupt handler, use it.No need to delete the
old bsd code, just to avoid disrupting the file too much. Make sure through
an assertion that it is never used, though.

Signed-off-by: Glauber Costa <glommer@cloudius-systems.com>

This version of the Xen interrupt handler tries to do as less work as possible
in the interrupt itself. The previous version and my previous fix attempt would
still clean the channels during interrupt.

Because now we have pending_sel still set in the irq thread, we can ditch
_irq_pending completely.

There is now only one xen_irq for the entire system, and therefore I am
registering one per cpu, since we will eventually have to process this in
different cpus. (for different event channels).

With this, in my (very course, host to guest) netperf test, I am achieving
9600 * 10^6 bps, while linux can reach ~10000 * 10^bps. So we're getting close:

Recv   Send    Send
Socket Socket  Message  Elapsed
Size   Size    Size     Time     Throughput
bytes  bytes   bytes    secs.    10^6bits/sec

 65536  16384  16384    10.00    9589.32

Signed-off-by: Glauber Costa <glommer@cloudius-systems.com>

Some of the fields in the xen shared structure need to be accessed atomically.
Move them to std::atomic so we can do that using C++11 primitives.

Signed-off-by: Glauber Costa <glommer@cloudius-systems.com>

Fix up the error message as suggested by Nadav.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

This adds a user-friendly error message to run.py if QEMU/KVM is not
installed.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

After 5f0e9733 htonl was called twice: once before calling
dhcp_mbuf::compose_request and once more inside this function.
Fix this, verified on wire packets with wireshark and audited code.

Signed-off-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Glauber Costa <glommer@cloudius-systems.com>

support etags

Some IP fields are stored in host order and some other in network order.
It makes discovery of the code difficult.
Convert all IP fields to host order and propagate to callers.

Signed-off-by: Benoit Canet <benoit@irqsave.net>

All the other fields of the pcpu structure BSD expects are initialized by
the event channel. Except for the cpu id, which the code expects to be already
initialized.

Signed-off-by: Glauber Costa <glommer@cloudius-systems.com>

Added comments to explain the target tags and TAGS,
also ignore the output file of ctags.

Signed-off-by: Yang Bai <hamo.by@gmail.com>

Update to latest mgmt.git:

* mgmt 97be48e...7a4db4e (1):
  > crash: Remove Rhono imports that are not used

It removes the last dependency to the old Rhino-based CLI...

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Some trivial comment cleanup and line-breaks in queue-mpsc.hh.

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Trivial fixes to comments in lock-free mutex implementation.

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Add a trivial sleep() test, which sleep()s for 2 seconds, and verifies
that this finishes and has slept for roughly 2 seconds.

I used this for debugging issue #26 - the attempts there ruined timers,
and in particular this trivial test hangs, as sleep() never returns.

(A note to our future automatic testing implementor: We need to allow
for the possibility that a test doesn't cleanly fail, but rather hangs,
and consider this a failure too).

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>

Switch to the mmu::page_size constant instead of inventing our own.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The 'offset' variable is first assigned to "bio->bio_offset" and then
immediately overwritten with "bio->bio_data & 0xfff". Fix that up.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Update to latest mgmt.git:

* mgmt 76f0e7e...97be48e (5):
  > md5sum: fix ugly spew when file doesn't exist
  > ls: handle non-existing files correctly
  > mgmt/crash: Fix printf format string in test.groovy
  > mgmt: Kill obsolete Rhino-based CLI code
  > mgmt/perf: Fix "perf callstack" usability gitch if there are no traces

The important part is Rhino-based CLI removal which seem to cause lot of
problems for people building OSv on other distributions. The usability
fixes to perf, ls, and md5sum commands are also worth the update.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Port 22 is usually taken on the host, so use port 2222 instead.

Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

add TAGS target to Makefile to support etags.

Signed-off-by: Yang Bai <hamo.by@gmail.com>

Having that implementation in the tree even though it was superceded is
confusing and really unnecessary.

Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

This regression test trigger issue #8
"Make Java InetAddress.getHostName() work" by exercising the DNS
resolver on localhost and a dns root server.

The test takes care of specifying NI_NOFQDN to resolve only the
hostname part of localhost ip.

It appears that the DNS ip is not communicated to the libc by
core/dhcp.cci: /etc/resolv.conf is not filled.

Test contributed while waiting for an fix idea to implement.

Signed-off-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

If Apache Ant is not found, the build fails with a cryptic error
message:

  [penberg@localhost osv]$ make
    ANT tests/bench
  Traceback (most recent call last):
    File "scripts/silentant.py", line 14, in <module>
      stderr = subprocess.PIPE)
    File "/usr/lib64/python2.7/subprocess.py", line 711, in __init__
      errread, errwrite)
    File "/usr/lib64/python2.7/subprocess.py", line 1308, in _execute_child
      raise child_exception
  OSError: [Errno 2] No such file or directory
  make: *** [all] Error 1

Make it more human friendly:

  [penberg@localhost osv]$ make
    ANT tests/bench
  Apache Ant not found. Please install the 'ant' package.
  make: *** [all] Error 1

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

OSerror exceptions throwed by subprocess.call are made silent using pass.
Fix this by properly catching the exception and pretty print it.

Signed-off-by: Benoit Canet <benoit@irqsave.net>

percpu had too little space allocated to support 64 vcpus, which
lead to a crash when booting with more than 13 vcpus. Fix it by
using a correct size to support 64 vcpus.

Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Avi Kivity <avi@cloudius-systems.com>

make[1]: Entering directory `/home/zwu/virt/osv/build/release'
  GEN gen/include/bits/alltypes.h
  AS arch/x64/boot16.o
  LD boot.bin
  AS arch/x64/boot.o
  CXX loader.o
In file included from <command-line>:0:0:
./../../compiler/include/intrinsics.hh:10:127: error: new declaration ‘unsigned char __builtin_ia32_addcarryx_u64(unsigned char, long unsigned int, long unsigned int, long long unsigned int*)’ [-Werror]
./../../compiler/include/intrinsics.hh:10:127: error: ambiguates built-in declaration ‘unsigned char __builtin_ia32_addcarryx_u64(unsigned char, long long unsigned int, long long unsigned int, long long unsigned int*)’ [-Werror]
cc1plus: all warnings being treated as errors
make[1]: *** [loader.o] Error 1
make[1]: Leaving directory `/home/zwu/virt/osv/build/release'
make: *** [all] Error 2

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>

Fix ant dependency on Debian

the ant package must also be explicitly specified, otherwise the following error will raise when building osv. 

raphaelsc@debian:~/Desktop/osv_dir/osv$ make
  ANT tests/bench
Traceback (most recent call last):
  File "scripts/silentant.py", line 14, in <module>
    stderr = subprocess.PIPE)
  File "/usr/lib/python2.6/subprocess.py", line 623, in __init__
    errread, errwrite)
  File "/usr/lib/python2.6/subprocess.py", line 1141, in _execute_child
    raise child_exception
OSError: [Errno 2] No such file or directory
make: *** [all] Error 1

Fix few minor but annoying coding style issues in drivers/virtio-blk.cc.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>