[PATCH 0/2] libvhost-user: lower dependency on QEMU headers

From: Marc-André Lureau Hi, libvhost-user is meant to be free of glib dependency, and easily copyable/reusable outside of QEMU. Clean-up some dependencies that crept in recently (the one remaining is qemu/atomic.h, from which a subset is used) Marc-André Lureau (2): libvhost-user: replace qem

[PATCH 1/2] libvhost-user: replace qemu/bswap.h with glibc endian.h

From: Marc-André Lureau Signed-off-by: Marc-André Lureau --- contrib/libvhost-user/libvhost-user.c | 77 ++- 1 file changed, 40 insertions(+), 37 deletions(-) diff --git a/contrib/libvhost-user/libvhost-user.c b/contrib/libvhost-user/libvhost-user.c index 5c73ffdd6b..1

[PATCH 2/2] libvhost-user: replace qemu/memfd.h usage

From: Marc-André Lureau Undo the damage from commit 5f9ff1eff3 ("libvhost-user: Support tracking inflight I/O in shared memory") which introduced glib dependency through osdep.h inclusion. libvhost-user.c tries to stay free from glib usage. Use glibc memfd_create directly when available (assume

[PATCH] dbus-vmstate: replace g_return with error handling

From: Marc-André Lureau Since g_input_stream_read_all() may return less than requested when the stream is malformed, we should treat this condition as a runtime user error (g_return are for programming errors). Signed-off-by: Marc-André Lureau --- backends/dbus-vmstate.c | 5 - 1 file chan

Re: [PATCH] meson: Fixes curses detection on msys2

On 16/11/20 22:31, Yonggang Luo wrote: Fixes regression on curses detection in msys2 by commit#925a40df2828d32d3aaaf022282cba81082fb263 Signed-off-by: Yonggang Luo --- meson.build | 13 + 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/meson.build b/meson.build inde

Re: [PATCH for-5.2] docs: Fix some typos (found by codespell)

On 17/11/20 20:34, Stefan Weil wrote: Fix also a similar typo in a code comment. Signed-off-by: Stefan Weil --- docs/can.txt | 8 docs/interop/vhost-user.rst | 2 +- docs/replay.txt | 2 +- docs/specs/ppc-spapr-numa.rst | 2 +- docs/system/depreca

[RFC 00/15] support subsets of bitmanip extension

From: Frank Chang This patchset implements RISC-V B-extension latest draft version (2020.10.26) Zbb, Zbs and Zba subset instructions. Specification: https://github.com/riscv/riscv-bitmanip/blob/master/bitmanip-draft.pdf The port is available here: https://github.com/sifive/qemu/tree/rvb-upstrea

Re: [PATCH for-5.2] meson: Fix build with --disable-guest-agent-msi

On 17/11/20 21:18, Stefan Weil wrote: The QGA MSI target requires several macros which are only available without --disable-guest-agent-msi. Don't define that target if configure was called with --disable-guest-agent-msi. Signed-off-by: Stefan Weil --- qga/meson.build | 36 ++

[RFC 01/15] target/riscv: reformat @sh format encoding for B-extension

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32.decode | 10 +- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode index 84080dd18ca..190ce469faf 100644 --- a/target/riscv/insn32.decode +++ b/target/ri

[RFC 05/15] target/riscv: rvb: pack two words into one register

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32-64.decode | 3 + target/riscv/insn32.decode | 3 + target/riscv/insn_trans/trans_rvb.c.inc | 30 target/riscv/translate.c| 92 + 4 files changed, 128 inse

[RFC 03/15] target/riscv: rvb: count bits set

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32-64.decode | 1 + target/riscv/insn32.decode | 1 + target/riscv/insn_trans/trans_rvb.c.inc | 12 target/riscv/translate.c| 21 + 4 files changed, 35 inser

[RFC 04/15] target/riscv: rvb: logic-with-negate

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32.decode | 4 +++ target/riscv/insn_trans/trans_rvb.c.inc | 18 ++ target/riscv/translate.c| 33 + 3 files changed, 55 insertions(+) diff --git a/target/riscv/insn

[RFC 02/15] target/riscv: rvb: count leading/trailing zeros

From: Kito Cheng Signed-off-by: Kito Cheng Signed-off-by: Frank Chang --- target/riscv/insn32-64.decode | 4 +++ target/riscv/insn32.decode | 7 +++- target/riscv/insn_trans/trans_rvb.c.inc | 47 target/riscv/translate.c| 48 +++

[RFC 09/15] target/riscv: rvb: shift ones

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32-64.decode | 4 ++ target/riscv/insn32.decode | 4 ++ target/riscv/insn_trans/trans_rvb.c.inc | 58 target/riscv/translate.c| 70 + 4 files cha

[RFC 06/15] target/riscv: rvb: min/max instructions

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32.decode | 4 target/riscv/insn_trans/trans_rvb.c.inc | 24 2 files changed, 28 insertions(+) diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode index 79aa40f03dd..372

[RFC 14/15] target/riscv: rvb: add/sub with postfix zero-extend

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32-64.decode | 7 + target/riscv/insn_trans/trans_rvb.c.inc | 38 + target/riscv/translate.c| 18 3 files changed, 63 insertions(+) diff --git a/target/riscv/insn

[RFC 08/15] target/riscv: rvb: single-bit instructions

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32-64.decode | 8 ++ target/riscv/insn32.decode | 9 ++ target/riscv/insn_trans/trans_rvb.c.inc | 90 ++ target/riscv/translate.c| 155 4 files change

[RFC 07/15] target/riscv: rvb: sign-extend instructions

From: Kito Cheng Signed-off-by: Kito Cheng --- target/riscv/insn32.decode | 2 ++ target/riscv/insn_trans/trans_rvb.c.inc | 13 + 2 files changed, 15 insertions(+) diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode index 372476f1d2c..626641333c6 1006

[RFC 10/15] target/riscv: rvb: rotate (left/right)

From: Kito Cheng Signed-off-by: Kito Cheng Signed-off-by: Frank Chang --- target/riscv/insn32-64.decode | 3 ++ target/riscv/insn32.decode | 3 ++ target/riscv/insn_trans/trans_rvb.c.inc | 54 + target/riscv/translate.c| 48 +

[PATCH v2 01/27] migration: Network Failover can't work with a paused guest

If we have a paused guest, it can't unplug the network VF device, so we wait there forever. Just change the code to give one error on that case. Signed-off-by: Juan Quintela --- migration/migration.c | 13 + 1 file changed, 13 insertions(+) diff --git a/migration/migration.c b/migr

[RFC 11/15] target/riscv: rvb: generalized reverse

From: Frank Chang Signed-off-by: Frank Chang --- target/riscv/bitmanip_helper.c | 79 + target/riscv/helper.h | 7 +++ target/riscv/insn32-64.decode | 2 + target/riscv/insn32.decode | 2 + target/riscv/insn_trans/tran

[RFC 12/15] target/riscv: rvb: generalized or-combine

From: Frank Chang Signed-off-by: Frank Chang --- target/riscv/bitmanip_helper.c | 49 + target/riscv/helper.h | 2 + target/riscv/insn32-64.decode | 2 + target/riscv/insn32.decode | 2 + target/riscv/insn_trans/trans_

[RFC 13/15] target/riscv: rvb: address calculation

From: Kito Cheng Signed-off-by: Kito Cheng Signed-off-by: Frank Chang --- target/riscv/insn32-64.decode | 3 +++ target/riscv/insn32.decode | 3 +++ target/riscv/insn_trans/trans_rvb.c.inc | 23 + target/riscv/translate.c| 33 +++

[PATCH v2 11/27] failover: remove standby_id variable

We can calculate it, and we only use it once anyways. Signed-off-by: Juan Quintela --- include/hw/virtio/virtio-net.h | 1 - hw/net/virtio-net.c| 11 +++ 2 files changed, 3 insertions(+), 9 deletions(-) diff --git a/include/hw/virtio/virtio-net.h b/include/hw/virtio/virtio-

Re: [PATCH for-6.0 1/6] qapi: Add query-accel command

Paolo, there's a question for you further down, marked with your name. Roman Bolshakov writes: > On Tue, Nov 17, 2020 at 09:51:58AM +0100, Markus Armbruster wrote: >> Eduardo Habkost writes: >> >> > On Mon, Nov 16, 2020 at 10:20:04AM -0600, Eric Blake wrote: >> >> On 11/16/20 7:10 AM, Roman Bo

[RFC 15/15] target/riscv: rvb: support and turn on B-extension from command line

From: Kito Cheng B-extension is default off, use cpu rv32 or rv64 with x-b=true to enable B-extension. Signed-off-by: Kito Cheng --- target/riscv/cpu.c | 4 target/riscv/cpu.h | 2 ++ 2 files changed, 6 insertions(+) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index 0bbfd7f4574.

[PATCH v2 02/27] failover: fix indentantion

Once there, remove not needed cast. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c| 33 +++-- softmmu/qdev-monitor.c | 4 ++-- 2 files changed, 17 insertions(+), 20 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 9179013ac4..1011a

[PATCH v2 14/27] failover: simplify virtio_net_find_primary()

a - is_my_primary() never sets one error b - If we return 1, primary_device_id is always set Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 18 +++--- 1 file changed, 3 insertions(+), 15 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 70fa372c08..88

[PATCH v2 00/27] Virtio net failover fixes

Hi This is a big rework of the network failover setup. General idea is: * We don't cache the name of the primary/standby devices We have several problems there with stale pointers * After this: - We can always remove either the primary/standby devices without trouble - Pluggin/unplugging works

[PATCH v2 05/27] failover: Remove unused parameter

Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 8 +++- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 786d313330..3f658d6246 100644 --- a/hw/net/virtio-net.c +++ b/hw/net/virtio-net.c @@ -855,9 +855,7 @@ static DeviceState

[PATCH v2 21/27] failover: make sure that id always exist

We check that it exist at device creation time, so we don't have to check anywhere else. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c| 3 --- softmmu/qdev-monitor.c | 4 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c inde

[PATCH v2 04/27] failover: primary bus is only used once, and where it is set

Just remove the struct member. Signed-off-by: Juan Quintela --- include/hw/virtio/virtio-net.h | 1 - hw/net/virtio-net.c| 8 2 files changed, 4 insertions(+), 5 deletions(-) diff --git a/include/hw/virtio/virtio-net.h b/include/hw/virtio/virtio-net.h index f4852ac27b..c8da

[PATCH v2 06/27] failover: Remove external partially_hotplugged property

It was only set "once", and with the wrong value. As far as I can see, libvirt still don't use it. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 4 1 file changed, 4 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 3f658d6246..6ca85627d8 100644 --- a/hw/ne

[PATCH v2 03/27] failover: Use always atomics for primary_should_be_hidden

Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 1011a524bf..a0fa63e7cb 100644 --- a/hw/net/virtio-net.c +++ b/hw/net/virtio-net.c @@ -3136,7 +3136,7 @@ static bool failover_rep

[PATCH v2 22/27] failover: remove failover_find_primary_device() error parameter

It can never give one error. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 9 +++-- 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index c708c03cf6..b994796734 100644 --- a/hw/net/virtio-net.c +++ b/hw/net/virtio-net.c @@ -832

[PATCH v2 08/27] failover: Rename bool to failover_primary_hidden

You should not use passive naming variables. And once there, be able to search for them. Signed-off-by: Juan Quintela --- include/hw/virtio/virtio-net.h | 3 ++- hw/net/virtio-net.c| 14 +++--- 2 files changed, 9 insertions(+), 8 deletions(-) diff --git a/include/hw/virtio/

[PATCH v2 07/27] failover: qdev_device_add() returns err or dev set

Never both. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 7 --- 1 file changed, 7 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 6ca85627d8..3e82108d42 100644 --- a/hw/net/virtio-net.c +++ b/hw/net/virtio-net.c @@ -803,13 +803,6 @@ static void failover_a

[PATCH v2 13/27] failover: Remove memory leak

Two things, at this point: * n->primary_device_id has to be set, otherwise virtio_net_find_primary don't work. So we have a leak here. * it has to be exactly the same that prim_dev->id because what qdev_find_recursive() does is just compare this two values. So remove the unneeded assignment

[PATCH v2 23/27] failover: split failover_find_primary_device_id()

So we can calculate the device id when we need it. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 63 + 1 file changed, 47 insertions(+), 16 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index b994796734..2c502c13fd 10064

[PATCH v2 10/27] failover: Remove primary_device_opts

It was really only used once, in failover_add_primary(). Just search for it on global opts when it is needed. Signed-off-by: Juan Quintela --- include/hw/virtio/virtio-net.h | 1 - hw/net/virtio-net.c| 21 + 2 files changed, 5 insertions(+), 17 deletions(-) dif

[PATCH v2 09/27] failover: g_strcmp0() knows how to handle NULL

Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index c221671852..e334f05352 100644 --- a/hw/net/virtio-net.c +++ b/hw/net/virtio-net.c @@ -818,7 +818,7 @@ static int is_my_primary(v

[PATCH v2 17/27] failover: virtio_net_connect_failover_devices() does nothing

It just calls virtio_net_find_primary(), so just update the callers. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 17 ++--- 1 file changed, 2 insertions(+), 15 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 747614ff2a..c6200b924e 100644 --- a/hw/

[PATCH v2 19/27] failover: simplify qdev_device_add() failover case

Just put allthe logic inside the same if. Signed-off-by: Juan Quintela --- softmmu/qdev-monitor.c | 15 --- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/softmmu/qdev-monitor.c b/softmmu/qdev-monitor.c index a25f5d612c..12b7540f17 100644 --- a/softmmu/qdev-monitor.c +

[PATCH v2 12/27] failover: Remove primary_device_dict

It was only used once. And we have there opts->id, so no need for it. Signed-off-by: Juan Quintela --- include/hw/virtio/virtio-net.h | 1 - hw/net/virtio-net.c| 17 - 2 files changed, 4 insertions(+), 14 deletions(-) diff --git a/include/hw/virtio/virtio-net.h b/i

Re: [PATCH v3] virtiofsd: Use --thread-pool-size=0 to mean no thread pool

On Tue, Nov 17, 2020 at 02:41:31PM -0500, Vivek Goyal wrote: > This is V3 of the patch. A minor change since V2 is to reverse the list > before executing requests. We are prepending elements to the list and that > means when we start processing requests, we are processing these in > the reverse ord

[PATCH v2 26/27] failover: simplify failover_unplug_primary

We can calculate device just once. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 25 ++--- 1 file changed, 10 insertions(+), 15 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index b37e9cd1d9..9203d81780 100644 --- a/hw/net/virtio-net.c +++ b/hw

[PATCH v2 24/27] failover: We don't need to cache primary_device_id anymore

Signed-off-by: Juan Quintela --- include/hw/virtio/virtio-net.h | 1 - hw/net/virtio-net.c| 20 ++-- 2 files changed, 10 insertions(+), 11 deletions(-) diff --git a/include/hw/virtio/virtio-net.h b/include/hw/virtio/virtio-net.h index fe353d8299..efef64e02f 100644 --

[PATCH v2 15/27] failover: should_be_hidden() should take a bool

We didn't use at all the -1 value, and we don't really care. It was only used for the cases when this is not the device that we are searching for. And in that case we should not hide the device. Once there, simplify virtio-Snet_primary_should_be_hidden. Signed-off-by: Juan Quintela --- includ

Re: [PATCH v2] net/e1000e_core: adjust count if RDH exceeds RDT in e1000e_ring_advance()

On Wed, Nov 18, 2020 at 4:56 AM Jason Wang wrote: > > > On 2020/11/13 下午6:31, Mauro Matteo Cascella wrote: > > The e1000e_write_packet_to_guest() function iterates over a set of > > receive descriptors by advancing rx descriptor head register (RDH) from > > its initial value to rx descriptor tail

[PATCH v2 16/27] failover: Rename function to hide_device()

You should not use pasive. Signed-off-by: Juan Quintela --- include/hw/qdev-core.h | 28 +++- hw/core/qdev.c | 4 ++-- hw/net/virtio-net.c| 7 +++ 3 files changed, 20 insertions(+), 19 deletions(-) diff --git a/include/hw/qdev-core.h b/include/hw/qdev-c

[PATCH v2 25/27] failover: Caller of this two functions already have primary_dev

Pass it as an argument. Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 27 ++- 1 file changed, 14 insertions(+), 13 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index 746ed3fb71..b37e9cd1d9 100644 --- a/hw/net/virtio-net.c +++ b/hw/net/virt

[PATCH for-5.2] s390x/pci: fix endianness issues

The zPCI group and function structures are big endian. However, we do not consistently store them as big endian locally, and are missing some conversions. Let's just store the structures as host endian instead and convert to big endian when actually handling the instructions retrieving the data.

[PATCH v2 18/27] failover: Rename to failover_find_primary_device()

This commit: * Rename them to failover_find_primary_devices() so - it starts with failover_ - it don't connect anything, just find the primary device * Create documentation for the function Signed-off-by: Juan Quintela --- hw/net/virtio-net.c | 12 +--- 1 file changed, 9 insertions(+

Re: [PATCH v2 00/27] Virtio net failover fixes

On Wed, Nov 18, 2020 at 09:37:21AM +0100, Juan Quintela wrote: > Hi > > This is a big rework of the network failover setup. General idea is: > * We don't cache the name of the primary/standby devices > We have several problems there with stale pointers > * After this: > - We can always remove e

[PATCH] hw/watchdog/wdt_diag288: Remove unnecessary includes

Both headers, sysbus.h and module.h, are not required to compile this file. Signed-off-by: Thomas Huth --- hw/watchdog/wdt_diag288.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/hw/watchdog/wdt_diag288.c b/hw/watchdog/wdt_diag288.c index 71a945f0bd..4c4b6a6ab7 100644 --- a/hw/watchdog/wd

[PATCH v2 20/27] failover: simplify qdev_device_add()

We don't need to walk the opts by hand. qmp_opt_get() already does that. And then we can remove the functions that did that walk. Signed-off-by: Juan Quintela --- softmmu/qdev-monitor.c | 32 ++-- 1 file changed, 6 insertions(+), 26 deletions(-) diff --git a/softmm

Re: [PATCH 0/2] libvhost-user: lower dependency on QEMU headers

On Wed, Nov 18, 2020 at 12:09:00PM +0400, marcandre.lur...@redhat.com wrote: > From: Marc-André Lureau > > Hi, > > libvhost-user is meant to be free of glib dependency, and easily > copyable/reusable outside of QEMU. Clean-up some dependencies that crept in > recently (the one remaining is qemu/

Re: [PATCH for-5.2] meson: Fix argument for makensis (build regression)

On 17/11/20 20:06, Stefan Weil wrote: `make installer` with a DLL directory was broken. Signed-off-by: Stefan Weil --- scripts/nsis.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/nsis.py b/scripts/nsis.py index e1c409344e..5135a05831 100644 --- a/scripts/nsis.

[PATCH v2 27/27] failover: Remove primary_dev member

Only three uses remained, and we can remove them on that case. Signed-off-by: Juan Quintela --- include/hw/virtio/virtio-net.h | 1 - hw/net/virtio-net.c| 55 +++--- 2 files changed, 24 insertions(+), 32 deletions(-) diff --git a/include/hw/virtio/virtio

Re: [PATCH] hvf: Fix segment selector format

On 16/11/20 21:04, Jessica Clarke wrote: The Requested Privilege Level field is 2 bits, the Table Indicator field is 1 bit and the Index field is the remaining 15 bits, with TI=0 meaning GDT and TI=1 meaning LDT. Signed-off-by: Jessica Clarke --- target/i386/hvf/x86.h | 8 1 file ch

Re: [PATCH 13/13] bcm2835_cprman: put some peripherals of bcm2835 cprman into the 'misc' category

Thomas Huth writes: > On 16/11/2020 18.00, Markus Armbruster wrote: >> Thomas Huth writes: >> >>> On 16/11/2020 14.25, Philippe Mathieu-Daudé wrote: Hi Gan, On 11/15/20 7:49 PM, Gan Qixin wrote: > Some peripherals of bcm2835 cprman have no category, put them into the > '

Re: Should bus-less devices default to .user_creatable = false? (was: [PATCH 13/13] bcm2835_cprman: put some peripherals of bcm2835 cprman into the 'misc' category)

On 18/11/2020 09.50, Markus Armbruster wrote: > Markus Armbruster writes: > > [...] >> qdev_device_add() looks like a bus-less device is usable if the machine >> provides a hotplug handler for it. Commit 03fcbd9dc5 "qdev: Check for >> the availability of a hotplug controller before adding a devi

[Bug 1903470] Re: qemu 5.1.0: Add UNIX socket support for netdev socket

With the fd= argument/property, you can setup a private socketpair connection already. Is this enough? -- You received this bug notification because you are a member of qemu- devel-ml, which is subscribed to QEMU. https://bugs.launchpad.net/bugs/1903470 Title: qemu 5.1.0: Add UNIX socket suppo

Re: [PATCH] hw/watchdog/wdt_diag288: Remove unnecessary includes

On 18.11.20 10:03, Thomas Huth wrote: > Both headers, sysbus.h and module.h, are not required to compile this file. > > Signed-off-by: Thomas Huth It is not a sysbus device and not a module. Reviewed-by: Christian Borntraeger > --- > hw/watchdog/wdt_diag288.c | 2 -- > 1 file changed, 2 d

Should bus-less devices default to .user_creatable = false? (was: [PATCH 13/13] bcm2835_cprman: put some peripherals of bcm2835 cprman into the 'misc' category)

Markus Armbruster writes: [...] > qdev_device_add() looks like a bus-less device is usable if the machine > provides a hotplug handler for it. Commit 03fcbd9dc5 "qdev: Check for > the availability of a hotplug controller before adding a device" seems > to be pertinent. Nope. A hotplug handler

Re: [PATCH for-5.2] s390x/pci: fix endianness issues

On Wed, 18 Nov 2020 09:51:54 +0100 Cornelia Huck wrote: [This is obviously a v2. Should not send patches before the first coffee :/] > The zPCI group and function structures are big endian. However, we do > not consistently store them as big endian locally, and are missing some > conversions. >

[PATCH 0/4] vhost-user: avoid g_return_val_if() in get/set_config()

Markus Armbruster pointed out that g_return_val_if() is meant for programming errors. It must not be used for input validation since it can be compiled out. Use explicit if statements instead. This patch series converts vhost-user device backends that use g_return_val_if() in get/set_config(). St

[PATCH 1/4] contrib/vhost-user-blk: avoid g_return_val_if() input validation

Do not validate input with g_return_val_if(). This API is intended for checking programming errors and is compiled out with -DG_DISABLE_CHECKS. Use an explicit if statement for input validation so it cannot accidentally be compiled out. Suggested-by: Markus Armbruster Signed-off-by: Stefan Hajno

[PATCH 2/4] contrib/vhost-user-gpu: avoid g_return_val_if() input validation

Do not validate input with g_return_val_if(). This API is intended for checking programming errors and is compiled out with -DG_DISABLE_CHECKS. Use an explicit if statement for input validation so it cannot accidentally be compiled out. Suggested-by: Markus Armbruster Signed-off-by: Stefan Hajno

[PATCH 3/4] contrib/vhost-user-input: avoid g_return_val_if() input validation

Do not validate input with g_return_val_if(). This API is intended for checking programming errors and is compiled out with -DG_DISABLE_CHECKS. Use an explicit if statement for input validation so it cannot accidentally be compiled out. Suggested-by: Markus Armbruster Signed-off-by: Stefan Hajno

[PATCH 4/4] block/export: avoid g_return_val_if() input validation

Do not validate input with g_return_val_if(). This API is intended for checking programming errors and is compiled out with -DG_DISABLE_CHECKS. Use an explicit if statement for input validation so it cannot accidentally be compiled out. Suggested-by: Markus Armbruster Signed-off-by: Stefan Hajno

Re: [PATCH for-6.0 1/6] qapi: Add query-accel command

On 18/11/20 09:36, Markus Armbruster wrote: The part that counts is do_configure_accelerator(). I works as follows: 1. Look up the chosen accelerator's QOM type (can fail) 2. Instantiate it (can't fail) 3. Set properties (can fail) 4. Connect the accelerator to the current machine (can fail)

Re: [PATCH for-5.2] s390x/pci: fix endianness issues

On 18/11/2020 09.51, Cornelia Huck wrote: > The zPCI group and function structures are big endian. However, we do > not consistently store them as big endian locally, and are missing some > conversions. > > Let's just store the structures as host endian instead and convert to > big endian when act

Re: [PATCH v1 2/6] tests: add prefixes to the bare mkdtemp calls

On 17/11/2020 18.36, Alex Bennée wrote: > The first step to debug a thing is to know what created the thing in > the first place. Add some prefixes so random tmpdir's have something > grep in the code. > > Signed-off-by: Alex Bennée > > --- > v2 > - fix long lines > --- > python/qemu/machine.

Re: [PATCH v1 4/6] gitlab: move remaining x86 check-tcg targets to gitlab

On 17/11/2020 18.36, Alex Bennée wrote: > The GCC check-tcg (user) test in particular was very prone to timing > out on Travis. We only actually need to move the some-softmmu builds > across as we already have coverage for linux-user. > > As --enable-debug-tcg does increase the run time somewhat a

Re: [PATCH 00/10] vhost/qemu: thread per IO SCSI vq

On Tue, Nov 17, 2020 at 01:13:14PM -0600, Mike Christie wrote: > On 11/17/20 10:40 AM, Stefan Hajnoczi wrote: > > On Thu, Nov 12, 2020 at 05:18:59PM -0600, Mike Christie wrote: > >> The following kernel patches were made over Michael's vhost branch: > >> > >> https://git.kernel.org/pub/scm/linux/ke

Re: [PATCH v1 6/6] gitlab-ci: Move trace backend tests across to gitlab

On 17/11/2020 18.36, Alex Bennée wrote: > From: Philippe Mathieu-Daudé > > Similarly to commit 8cdb2cef3f1, move the trace backend > tests to GitLab. > > Note the User-Space Tracer backend is still tested on > Ubuntu by the s390x jobs on Travis-CI. > > Signed-off-by: Philippe Mathieu-Daudé > M

Re: Property '.hmat' not found

> > Could you check what happens if you use > "-machine pc-i440fx-5.2,hmat=on" instead of "-machine hmat=on"? It failed, but `-machine pc-i440fx-4.2` worked without `hmat` option. So, the problem could be that I didn't fully compile QEMU, which specified `--target-list=x86_64-softmmu` in the conf

[PATCH] cirrus-ci: parallelize "make check"

Otherwise, the tests risk failing due to timeouts. Signed-off-by: Paolo Bonzini --- .cirrus.yml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.cirrus.yml b/.cirrus.yml index f0209b7a3e..111e19f1bf 100644 --- a/.cirrus.yml +++ b/.cirrus.yml @@ -29,7 +29,7 @@ macos_task:

[PULL 0/7] HVF and Windows changes for QEMU 5.2-rc

The following changes since commit 66a300a107ec286725bdc943601cbd4247b82158: Update version for v5.2.0-rc2 release (2020-11-17 22:58:10 +) are available in the Git repository at: https://gitlab.com/bonzini/qemu.git tags/for-upstream for you to fetch changes up to 29cf16db23e605769f6d89c

[PULL 3/7] meson: Fix build with --disable-guest-agent-msi

From: Stefan Weil The QGA MSI target requires several macros which are only available without --disable-guest-agent-msi. Don't define that target if configure was called with --disable-guest-agent-msi. Signed-off-by: Stefan Weil Message-Id: <20201117201834.408892-1...@weilnetz.de> Signed-off-b

[PULL 7/7] buildsys: Help git-diff adding .gitattributes config file

From: Philippe Mathieu-Daudé Since commits 0979ed017f0 ("meson: rename .inc.h files to .h.inc") and 139c1837db7 ("meson: rename included C source files to .c.inc") 'git-diff --function-context' stopped displaying C function context correctly. We can help git-diff by providing attributes to the .[

[PULL 1/7] meson: Fix argument for makensis (build regression)

From: Stefan Weil `make installer` with a DLL directory was broken. Signed-off-by: Stefan Weil Message-Id: <20201117190640.390359-1...@weilnetz.de> Reviewed-by: Marc-André Lureau Signed-off-by: Paolo Bonzini --- scripts/nsis.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --gi

[PULL 2/7] docs: Fix some typos (found by codespell)

From: Stefan Weil Fix also a similar typo in a code comment. Signed-off-by: Stefan Weil Message-Id: <20201117193448.393472-1...@weilnetz.de> Reviewed-by: Michael S. Tsirkin Signed-off-by: Paolo Bonzini --- docs/can.txt | 8 docs/interop/vhost-user.rst | 2 +- docs

[PULL 4/7] meson: Fixes curses detection on msys2

From: Yonggang Luo Fixes regression on curses detection in msys2 by commit#925a40df2828d32d3aaaf022282cba81082fb263 Signed-off-by: Yonggang Luo Message-Id: <20201116213106.589-1-luoyongg...@gmail.com> Signed-off-by: Paolo Bonzini --- meson.build | 13 + 1 file changed, 9 insertion

[PULL 5/7] hvf: Gate RDTSCP on CPU_BASED2_RDTSCP, not just CPU_BASED_TSC_OFFSET

From: Jessica Clarke Buglink: https://bugs.launchpad.net/qemu/+bug/1894836 Signed-off-by: Jessica Clarke Message-Id: <20201116200319.28138-1-jrt...@jrtc27.com> Signed-off-by: Paolo Bonzini --- target/i386/hvf/x86_cpuid.c | 4 1 file changed, 4 insertions(+) diff --git a/target/i386/hvf/x

[PULL 6/7] hvf: Fix segment selector format

From: Jessica Clarke The Requested Privilege Level field is 2 bits, the Table Indicator field is 1 bit and the Index field is the remaining 15 bits, with TI=0 meaning GDT and TI=1 meaning LDT. Signed-off-by: Jessica Clarke Message-Id: <20201116200414.28286-1-jrt...@jrtc27.com> Signed-off-by: Pa

Re: [PATCH for-5.2] s390x/pci: fix endianness issues

On Wed, 18 Nov 2020 10:38:00 +0100 Thomas Huth wrote: > On 18/11/2020 09.51, Cornelia Huck wrote: > > The zPCI group and function structures are big endian. However, we do > > not consistently store them as big endian locally, and are missing some > > conversions. > > > > Let's just store the st

[Bug 1163034] Re: linux-user mode can't handle guest setting a very small RLIMIT_AS (hangs running gnutls28, coreutils configure check code)

It's been a while since the last change to this ticket ... Has this ever been implemented? -- You received this bug notification because you are a member of qemu- devel-ml, which is subscribed to QEMU. https://bugs.launchpad.net/bugs/1163034 Title: linux-user mode can't handle guest setting a

[Bug 1243968] Re: VMware ESXi on QEmu Kernel Panic

The QEMU project is currently considering to move its bug tracking to another system. For this we need to know which bugs are still valid and which could be closed already. Thus we are setting older bugs to "Incomplete" now. If you still think this bug report here is valid, then please switch the

[RFC v3 2/9] i386: move whpx accel files into whpx/

Signed-off-by: Claudio Fontana --- MAINTAINERS | 5 + target/i386/meson.build | 5 + target/i386/whpx/meson.build | 4 target/i386/{ => whpx}/whp-dispatch.h | 0 target/i386/{ => whpx}/whpx-all.c | 0 target/i386/{ => whpx}/whpx-cp

[RFC v3 0/9] i386 cleanup

Hi all, in this cleanup series we are back at RFC, and back at providing the whole series together. The last patch in particular contains a different way to provide the accelerator-specific i386 CPU functionality, compared with RFCv1. In this series I start by moving all i386-specific accelerator

[RFC v3 4/9] i386: hvf: remove stale MAINTAINERS entry for old hvf stubs

Signed-off-by: Claudio Fontana --- MAINTAINERS | 1 - 1 file changed, 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 463317784c..e892dd2220 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -443,7 +443,6 @@ M: Cameron Esfahani M: Roman Bolshakov W: https://wiki.qemu.org/Features/HVF

[RFC v3 3/9] i386: move hax accel files into hax/

Signed-off-by: Claudio Fontana --- MAINTAINERS | 2 +- target/i386/{ => hax}/hax-all.c | 0 target/i386/{ => hax}/hax-cpus.c | 0 target/i386/{ => hax}/hax-cpus.h | 0 target/i386/{ => hax}/hax-i386.h | 6 +++--- target/i386/{ => hax}/hax-interface.h

[RFC v3 6/9] i386: move cpu dump out of helper.c into cpu-dump.c

Signed-off-by: Claudio Fontana --- target/i386/cpu-dump.c | 538 target/i386/cpu.h | 1 + target/i386/helper.c| 514 -- target/i386/meson.build | 1 + 4 files changed, 540 insertions(+), 514 deletions(-)

[RFC v3 5/9] i386: move TCG accel files into tcg/

Signed-off-by: Claudio Fontana --- target/i386/meson.build | 14 +- target/i386/{ => tcg}/bpt_helper.c | 0 target/i386/{ => tcg}/cc_helper.c | 0 target/i386/{ => tcg}/excp_helper.c | 0 target/i386/{ => tcg}/fpu_helper.c | 0 target/i386/{ => tcg}/int_helper.c |

[RFC v3 1/9] i386: move kvm accel files into kvm/

Signed-off-by: Claudio Fontana --- MAINTAINERS | 2 +- hw/i386/fw_cfg.c | 2 +- hw/i386/intel_iommu.c| 2 +- hw/i386/kvm/apic.c | 2 +- hw/i386/kvm/clock.c | 2 +- hw/i386/microvm.c

[RFC v3 9/9] i386: split cpu accelerators from cpu.c

split cpu.c into: cpu.ccpuid and common x86 cpu functionality host-cpu.c host x86 cpu functions and "host" cpu type kvm/cpu.cKVM x86 cpu type hvf/cpu.cHVF x86 cpu type tcg/cpu.cTCG x86 cpu type The accel interface of the X86CPUClass is set at MODULE_INIT_

[Bug 1565395] Re: qemu-2.4.1 fails when compiled against pulseaudio

The QEMU project is currently considering to move its bug tracking to another system. For this we need to know which bugs are still valid and which could be closed already. Thus we are setting older bugs to "Incomplete" now. If you still think this bug report here is valid, then please switch the

Re: About 'qemu-security' mailing list

Hello Dan, Stefan, +-- On Tue, 17 Nov 2020, Daniel P. Berrangé wrote --+ | On Tue, Nov 17, 2020 at 04:19:42PM +, Stefan Hajnoczi wrote: | > Dan and I tried out confidential issues and unfortunately it is | > currently too limited for our workflow. | > | > It is not possible to add non-memb

  1   2   3   4   >