Clang v7.0.1 does not like the __int128 variable type for inline
assembly on s390x:
In file included from fpu/softfloat.c:97:
include/fpu/softfloat-macros.h:647:9: error: inline asm error:
This value type register class is not natively supported!
asm("dlgr %0, %1" : "+r"(n) : "r"(d));
^
Disable this code part there now when compiling with Clang, so that
the generic code gets used instead.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
The added branch to the FMA ops is marked as unlikely and therefore
its impact on performance (measured with fp-bench) is within noise range
when measured on an Intel(R) Xeon(R) Gold 6142 CPU @ 2.60GHz.
Reported-by: Laurent Desnogues <laurent.desnogues@gmail.com>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
We get HOST_WORDS_BIGENDIAN from config-host.h, but the include
is missing. Fix it.
This fixes `make check-softfloat' on big endian hosts.
Reported-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
At this point random_ops[] only contains normals, so there's
no need to do anything to them. In fact, raising the exponent
here can make the output !normal, which is precisely
what the comment says we want to avoid.
Signed-off-by: Emilio G. Cota <cota@braap.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
The second test in the branches is wrong; fix while converting
to a switch statement, which is easier to get right.
Signed-off-by: Emilio G. Cota <cota@braap.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
- Get rid of global_qtests in some more qtests
- typedef cleanups
- Fixes for compiling with Clang
- Force C standard to gnu99
- Don't use -nographic in qtests
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)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=X/Yb
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2019-01-22' into staging
- Use more CONFIG Makefile switches for qtests
- Get rid of global_qtests in some more qtests
- typedef cleanups
- Fixes for compiling with Clang
- Force C standard to gnu99
- Don't use -nographic in qtests
# gpg: Signature made Tue 22 Jan 2019 06:18:41 GMT
# gpg: using RSA key 2ED9D774FE702DB5
# gpg: Good signature from "Thomas Huth <th.huth@gmx.de>"
# gpg: aka "Thomas Huth <thuth@redhat.com>"
# gpg: aka "Thomas Huth <huth@tuxfamily.org>"
# gpg: aka "Thomas Huth <th.huth@posteo.de>"
# Primary key fingerprint: 27B8 8847 EEE0 2501 18F3 EAB9 2ED9 D774 FE70 2DB5
* remotes/huth-gitlab/tags/pull-request-2019-01-22: (26 commits)
tests: remove rule for nonexisting qdev-monitor-test
tests/hexloader-test: Don't pass -nographic to the QEMU under test
configure: Force the C standard to gnu99
seccomp: Work-around GCC 4.x bug in gnu99 mode
block: Work-around a bug in libiscsi 1.9.0 when used in gnu99 mode
linux-user: Fix compilation with clang 3.4
virtio-net: Fix VirtIONet typedef redefinition
ppc: Fix duplicated typedefs to be able to compile with Clang in gnu99 mode
ppc: Move spapr-related prototypes from xics.h into a seperate header file
ui/console: Remove PixelFormat from qemu/typedefs.h
ui/console: Remove MouseTransformInfo from qemu/typedefs.h
ui/console: Remove DisplayState/DisplaySurface from "qemu/typedefs.h"
ui/console: Remove QemuDmaBuf from "qemu/typedefs.h"
audio: Remove AudioState from "qemu/typedefs.h"
hw/i386: Remove PCMachineClass from "qemu/typedefs.h"
hw/char/serial: Remove SerialState from "qemu/typedefs.h"
hw/bt: Remove HCIInfo from "qemu/typedefs.h"
hw/i2c/smbus: Remove SMBusDevice from "qemu/typedefs.h"
hw/ide/ahci: Remove AllwinnerAHCIState from "qemu/typedefs.h"
hw/pcmcia: Remove PCMCIACardState from "qemu/typedefs.h"
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This test was merged into drive_del-test in 2014.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Fixes: e2f3f22188 ("Merge of qdev-monitor-test, blockdev-test")
Signed-off-by: Thomas Huth <thuth@redhat.com>
The hexloader test invokes QEMU with the -nographic argument. This
is unnecessary, because the qtest_initf() function will pass it
-display none, which suffices to disable the graphical window.
It also means that the QEMU process will make the stdin/stdout
O_NONBLOCK. Since O_NONBLOCK is not per-file descriptor but per
"file description", this non-blocking behaviour is then shared
with any other process that's using the stdin/stdout of the
'make check' run, including make itself. This can result in make
falling over with "make: write error: stdout" because it got
an unexpected EINTR trying to write output messages to the terminal.
This is particularly noticable if running 'make check' in a loop with
while make check; do true; done
(It does not affect single make check runs so much because the
shell will remove the O_NONBLOCK status before it reads the
terminal for interactive input.)
Remove the unwanted -nographic argument.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Different versions of GCC and Clang use different versions of the C standard.
This repeatedly caused problems already, e.g. with duplicated typedefs:
https://lists.gnu.org/archive/html/qemu-devel/2018-11/msg05829.html
or with for-loop variable initializers:
https://lists.gnu.org/archive/html/qemu-devel/2019-01/msg00237.html
To avoid these problems, we should enforce the C language version to the
same level for all compilers. Since our minimum compiler versions is
GCC v4.8, our best option is "gnu99" for C code right now ("gnu17" is not
available there yet, and "gnu11" is marked as "experimental"), and "gnu++98"
for the few C++ code that we have in the repository.
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
We'd like to compile QEMU with -std=gnu99, but GCC 4.8 currently
fails to compile qemu-seccomp.c in this mode:
qemu-seccomp.c:45:1: error: initializer element is not constant
};
^
qemu-seccomp.c:45:1: error: (near initialization for ‘sched_setscheduler_arg[0]’)
This is due to a compiler bug which has just been fixed in GCC 5.0:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63567
Since we still want to support GCC 4.8 for a while and also want to use
gnu99 mode, work-around the issue by expanding the macro manually.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Eduardo Otubo <otubo@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
The header "scsi-lowlevel.h" of libiscsi 1.9.0 contains some bad
"inline" prototype definitions which GCC refuses to compile in its
gnu99 mode:
In file included from block/iscsi.c:52:0:
/usr/include/iscsi/scsi-lowlevel.h:810:13: error: inline function
‘scsi_set_uint16’ declared but never defined [-Werror]
inline void scsi_set_uint16(unsigned char *c, uint16_t val);
^
/usr/include/iscsi/scsi-lowlevel.h:809:13: error: inline function
‘scsi_set_uint32’ declared but never defined [-Werror]
inline void scsi_set_uint32(unsigned char *c, uint32_t val);
^
[...]
This has been fixed by upstream libiscsi in version 1.10.0 (see
https://github.com/sahlberg/libiscsi/commit/7692027d6c11 ), but
since we still want to support 1.9.0 for CentOS 7 / RHEL7, we
have to work-around the issue by redefining the "inline" keyword
to use the old "gnu89" mode behavior via "gnu_inline" instead.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Clang version 3.4.2 does not know the -Wpragmas option yet and bails
out with an error when we try to disable it in linux-user/qemu.h.
Fortunately, clang has a __has_warning() macro which allows us to add
an explicit check for the option that we want to ignore. With that we
can check for the availability of "-Waddress-of-packed-member" properly
and do not need the "-Wpragmas" at all here.
Fixes: 850d5e330a
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Commit 2974e916df introduced the VirtioNetRscChain structure which
refer to a VirtIONet, declared later, thus required VirtIONet typedef
to use a forward declaration.
However, when compiling with Clang in -std=gnu99 mode, this triggers
the following warning/error:
CC hw/net/virtio-net.o
In file included from qemu/hw/net/virtio-net.c:22:
include/hw/virtio/virtio-net.h:189:3: error: redefinition of typedef 'VirtIONet' is a C11 feature [-Werror,-Wtypedef-redefinition]
} VirtIONet;
^
include/hw/virtio/virtio-net.h:110:26: note: previous definition is here
typedef struct VirtIONet VirtIONet;
^
1 error generated.
make: *** [rules.mak:69: hw/net/virtio-net.o] Error 1
Fix it by removing the duplicate typedef definition.
Reported-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
When compiling the ppc code with clang and -std=gnu99, there are a
couple of warnings/errors like this one:
CC ppc64-softmmu/hw/intc/xics.o
In file included from hw/intc/xics.c:35:
include/hw/ppc/xics.h:43:25: error: redefinition of typedef 'ICPState' is a C11 feature
[-Werror,-Wtypedef-redefinition]
typedef struct ICPState ICPState;
^
target/ppc/cpu.h:1181:25: note: previous definition is here
typedef struct ICPState ICPState;
^
Work around the problems by including the proper headers in spapr.h
and by using struct forward declarations in cpu.h.
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
When compiling with Clang in -std=gnu99 mode, there is a warning/error:
CC ppc64-softmmu/hw/intc/xics_spapr.o
In file included from /home/thuth/devel/qemu/hw/intc/xics_spapr.c:34:
/home/thuth/devel/qemu/include/hw/ppc/xics.h:203:34: error: redefinition of typedef 'sPAPRMachineState' is a C11 feature
[-Werror,-Wtypedef-redefinition]
typedef struct sPAPRMachineState sPAPRMachineState;
^
/home/thuth/devel/qemu/include/hw/ppc/spapr_irq.h:25:34: note: previous definition is here
typedef struct sPAPRMachineState sPAPRMachineState;
^
We have to remove the duplicated typedef here and include "spapr.h" instead.
But "spapr.h" should not be included for the pnv machine files. So move
the spapr-related prototypes into a new file called "xics_spapr.h" instead.
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Header files requiring PixelFormat already include "ui/qemu-pixman.h".
To clean "qemu/typedefs.h", move the declaration to "ui/qemu-pixman.h"
(removing the forward declaration).
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Header files requiring MouseTransformInfo already include "ui/console.h".
To clean "qemu/typedefs.h", move the declaration to "ui/console.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring DisplayState/DisplaySurface already include "ui/console.h".
To clean "qemu/typedefs.h", move the declarations to "ui/console.h"
(removing DisplaySurface forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring QemuDmaBuf already include "ui/console.h".
To clean "qemu/typedefs.h", move the declaration to "ui/console.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring AudioState already include "audio_int.h".
To clean "qemu/typedefs.h", move the declaration to "audio_int.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring PCMachineClass already include "hw/i386/pc.h".
To clean "qemu/typedefs.h", move the declaration to "hw/i386/pc.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring SerialState already include "hw/char/serial.h".
To clean "qemu/typedefs.h", move the declaration to "hw/char/serial.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring HCIInfo already include "sysemu/bt.h".
To clean "qemu/typedefs.h", move the declaration to "sysemu/bt.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring SMBusDevice already include "hw/i2c/smbus.h".
To clean "qemu/typedefs.h", move the forward declaration
to "hw/i2c/smbus.h".
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Files requiring AllwinnerAHCIState already include "hw/ide/ahci.h".
To clean "qemu/typedefs.h", move the declaration to "hw/ide/ahci.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
There is only one header file requiring this typedef (hw/arm/pxa.h),
let it include "hw/pcmcia.h" directly to simplify "qemu/typedefs.h".
To clean "qemu/typedefs.h", move the declaration to "hw/pcmcia.h"
(removing the forward declaration).
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
[thuth: slightly tweaked commit message]
Signed-off-by: Thomas Huth <thuth@redhat.com>
PS2State is only used in "hw/input/ps2.h", there is no
need to expose it via "qemu/typedefs.h".
To clean "qemu/typedefs.h", move the forward declaration
to "hw/input/ps2.h".
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Pass around the QTestState, so that we can finally get rid of the
out-of-favor global_qtest variable in this file, too.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Pass around the QTestState from function to function, so that we can finally
get rid of the out-of-favor global_qtest variable in this file, too.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Pass around the test state explicitly, to be able to use the qtest_in*()
and qtest_out*() function in this test.
Signed-off-by: Thomas Huth <thuth@redhat.com>
To be able to build and test QEMU binaries where certain devices or machines
are disabled, we have to use the right CONFIG_* switches to run certain tests
only if the corresponding device or machine really has been compiled into
the binary.
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
To be able to build and test QEMU binaries where certain devices are
disabled, we have to use the right CONFIG_* switches to run certain
tests only if the corresponding device really has been compiled into
the binary.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Any good new feature deserves some regression testing :)
Coverage includes:
- 223: what happens when there are 0 or more than 1 export,
proof that we can see multiple contexts including qemu:dirty-bitmap
- 233: proof that we can list over TLS, and that mix-and-match of
plain/TLS listings will behave sanely
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-22-eblake@redhat.com>
Commit 3d068aff forgot to advertise available qemu: contexts
when the client requests a list with 0 queries. Furthermore,
3.0 shipped with a qemu-img hack of x-dirty-bitmap (commit
216ee365) that _silently_ acts as though the entire image is
clean if a requested bitmap is not present. Both bugs have
been recently fixed, so that a modern qemu server gives full
context output right away, and the client refuses a
connection if a requested x-dirty-bitmap was not found.
Still, it is likely that there will be users that have to
work with a mix of old and new qemu versions, depending on
which features get backported where, at which point being
able to rely on 'qemu-img --list' output to know for sure
whether a given NBD export has the desired dirty bitmap is
much nicer than blindly connecting and risking that the
entire image may appear clean. We can make our --list code
smart enough to work around buggy servers by tracking
whether we've seen any qemu: replies in the original 0-query
list; if not, repeat with a single query on "qemu:" (which
may still have no replies, but then we know for sure we
didn't trip up on the server bug).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-21-eblake@redhat.com>
We want to be able to detect whether a given qemu NBD server is
exposing the right export(s) and dirty bitmaps, at least for
regression testing. We could use 'nbd-client -l' from the upstream
NBD project to list exports, but it's annoying to rely on
out-of-tree binaries; furthermore, nbd-client doesn't necessarily
know about all of the qemu NBD extensions. Thus, it is time to add
a new mode to qemu-nbd that merely sniffs all possible information
from the server during handshake phase, then disconnects and dumps
the information.
This patch actually implements --list/-L, while reusing other
options such as --tls-creds for now designating how to connect
as the client (rather than their non-list usage of how to operate
as the server).
I debated about adding this functionality to something akin to
'qemu-img info' - but that tool does not readily lend itself
to connecting to an arbitrary NBD server without also tying to
a specific export (I may, however, still add ImageInfoSpecificNBD
for reporting the bitmaps available when connecting to a single
export). And, while it may feel a bit odd that normally
qemu-nbd is a server but 'qemu-nbd -L' is a client, we are not
really making the qemu-nbd binary that much larger, because
'qemu-nbd -c' has to operate as both server and client
simultaneously across two threads when feeding the kernel module
for /dev/nbdN access.
Sample output:
$ qemu-nbd -L
exports available: 1
export: ''
size: 65536
flags: 0x4ed ( flush fua trim zeroes df cache )
min block: 512
opt block: 4096
max block: 33554432
available meta contexts: 1
base:allocation
Note that the output only lists sizes if the server sent
NBD_FLAG_HAS_FLAGS, because a newstyle server does not give
the size otherwise. It has the side effect that for really
old servers that did not send any flags, the size is not
output even though it was available. However, I'm not too
concerned about that - oldstyle servers are (rightfully)
getting less common to encounter (qemu 3.0 was the last
version where we even serve it), and most existing servers
that still even offer oldstyle negotiation (such as nbdkit)
still send flags (since that was added to the NBD protocol
in 2007 to permit read-only connections).
Not done here, but maybe worth future experiments: capture
the meat of NBDExportInfo into a QAPI struct, and use the
generated QAPI pretty-printers instead of hand-rolling our
output loop. It would also permit us to add a JSON output
mode for machine parsing.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Message-Id: <20190117193658.16413-20-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
We want to be able to detect whether a given qemu NBD server is
exposing the right export(s) and dirty bitmaps, at least for
regression testing. We could use 'nbd-client -l' from the upstream
NBD project to list exports, but it's annoying to rely on
out-of-tree binaries; furthermore, nbd-client doesn't necessarily
know about all of the qemu NBD extensions. Thus, we plan on adding
a new mode to qemu-nbd that merely sniffs all possible information
from the server during handshake phase, then disconnects and dumps
the information.
This patch continues the work of the previous patch, by adding the
ability to track the list of available meta contexts into
NBDExportInfo. It benefits from the recent refactoring patches
with a new nbd_list_meta_contexts() that reuses much of the same
framework as setting a meta context.
Note: a malicious server could exhaust memory of a client by feeding
an unending loop of contexts; perhaps we could place a limit on how
many we are willing to receive. But this is no different from our
earlier analysis on a server sending an unending list of exports,
and the death of a client due to memory exhaustion when the client
was going to exit soon anyways is not really a denial of service
attack.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-19-eblake@redhat.com>
We want to be able to detect whether a given qemu NBD server is
exposing the right export(s) and dirty bitmaps, at least for
regression testing. We could use 'nbd-client -l' from the upstream
NBD project to list exports, but it's annoying to rely on
out-of-tree binaries; furthermore, nbd-client doesn't necessarily
know about all of the qemu NBD extensions. Thus, we plan on adding
a new mode to qemu-nbd that merely sniffs all possible information
from the server during handshake phase, then disconnects and dumps
the information.
This patch adds the low-level client code for grabbing the list
of exports. It benefits from the recent refactoring patches, in
order to share as much code as possible when it comes to doing
validation of server replies. The resulting information is stored
in an array of NBDExportInfo which has been expanded to any
description string, along with a convenience function for freeing
the list.
Note: a malicious server could exhaust memory of a client by feeding
an unending loop of exports; perhaps we should place a limit on how
many we are willing to receive. But note that a server could
reasonably be serving an export for every file in a large directory,
where an arbitrary limit in the client means we can't list anything
from such a server; the same happens if we just run until the client
fails to malloc() and thus dies by an abort(), where the limit is
no longer arbitrary but determined by available memory. Since the
client is already planning on being short-lived, it's hard to call
this a denial of service attack that would starve off other uses,
so it does not appear to be a security issue.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Message-Id: <20190117193658.16413-18-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Rename the function to nbd_opt_info_or_go() with an added parameter
and slight changes to comments and trace messages, in order to
reuse the function for NBD_OPT_INFO.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20190117193658.16413-17-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Another refactoring creating nbd_negotiate_finish_oldstyle()
for further reuse during 'qemu-nbd --list'.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Message-Id: <20190117193658.16413-16-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
An upcoming patch will add the ability for qemu-nbd to list
the services provided by an NBD server. Share the common
code of the TLS handshake by splitting the initial exchange
into a separate function, leaving only the export handling
in the original function. Functionally, there should be no
change in behavior in this patch, although some of the code
motion may be difficult to follow due to indentation changes
(view with 'git diff -w' for a smaller changeset).
I considered an enum for the return code coordinating state
between the two functions, but in the end just settled with
ample comments.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-15-eblake@redhat.com>
The function could only ever return 0 or -EINVAL; make this
clearer by dropping a useless 'fail:' label.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-14-eblake@redhat.com>
Extract portions of nbd_negotiate_simple_meta_context() to
a new function nbd_receive_one_meta_context() that copies the
pattern of nbd_receive_list() for performing the argument
validation of one reply. The error message when the server
replies with more than one context changes slightly, but
that shouldn't happen in the common case.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-13-eblake@redhat.com>
Refactor nbd_negotiate_simple_meta_context() to pull out the
code that can be reused to send a LIST request for 0 or 1 query.
No semantic change. The old comment about 'sizeof(uint32_t)'
being equivalent to '/* number of queries */' is no longer
needed, now that we are computing 'sizeof(queries)' instead.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Message-Id: <20190117193658.16413-12-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Pass 'info' instead of three separate parameters related to info,
when requesting the server to set the meta context. Update the
NBDExportInfo struct to rename the received id field to match the
fact that we are currently overloading the field to match whatever
context the user supplied through the x-dirty-bitmap hack, as well
as adding a TODO comment to remind future patches about a desire
to request two contexts at once.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-11-eblake@redhat.com>
Refactor the 'name' parameter of nbd_receive_negotiate() from
being a separate parameter into being part of the in-out 'info'.
This also spills over to a simplification of nbd_opt_go().
The main driver for this refactoring is that an upcoming patch
would like to add support to qemu-nbd to list information about
all exports available on a server, where the name(s) will be
provided by the server instead of the client. But another benefit
is that we can now allow the client to explicitly specify the
empty export name "" even when connecting to an oldstyle server
(even if qemu is no longer such a server after commit 7f7dfe2a).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-10-eblake@redhat.com>
Right now, nbd_receive_list() is only called by
nbd_receive_query_exports(), which in turn is only called if the
server lacks NBD_OPT_GO but has working option negotiation, and is
merely used as a quality-of-implementation trick since servers
can't give decent errors for NBD_OPT_EXPORT_NAME. However, servers
that lack NBD_OPT_GO are becoming increasingly rare (nbdkit was a
latecomer, in Aug 2018, but qemu has been such a server since commit
f37708f6 in July 2017 and released in 2.10), so it no longer makes
sense to micro-optimize that function for performance.
Furthermore, when debugging a server's implementation, tracing the
full reply (both names and descriptions) is useful, not to mention
that upcoming patches adding 'qemu-nbd --list' will want to collect
that data. And when you consider that a server can send an export
name up to the NBD protocol length limit of 4k; but our current
NBD_MAX_NAME_SIZE is only 256, we can't trace all valid server
names without more storage, but 4k is large enough that the heap
is better than the stack for long names.
Thus, I'm changing the division of labor, with nbd_receive_list()
now always malloc'ing a result on success (the malloc is bounded
by the fact that we reject servers with a reply length larger
than 32M), and moving the comparison to 'wantname' to the caller.
There is a minor change in behavior where a server with 0 exports
(an immediate NBD_REP_ACK reply) is now no longer distinguished
from a server without LIST support (NBD_REP_ERR_UNSUP); this
information could be preserved with a complication to the calling
contract to provide a bit more information, but I didn't see the
point. After all, the worst that can happen if our guess at a
match is wrong is that the caller will get a cryptic disconnect
when NBD_OPT_EXPORT_NAME fails (which is no different from what
would happen if we had not tried LIST), while treating an empty
list as immediate failure would prevent connecting to really old
servers that really did lack LIST. Besides, NBD servers with 0
exports are rare (qemu can do it when using QMP nbd-server-start
without nbd-server-add - but qemu understands NBD_OPT_GO and
thus won't tickle this change in behavior).
Fix the spelling of foundExport to match coding standards while
in the area.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20190117193658.16413-9-eblake@redhat.com>