aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2023-01-13hw/mips/malta: Merge common BL code as bl_setup_gt64120_jump_kernel()Philippe Mathieu-Daudé
Merge common code shared between write_bootloader() and write_bootloader_nanomips() into bl_setup_gt64120_jump_kernel(). Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-12-philmd@linaro.org>
2023-01-13hw/mips/malta: Use bootloader generator API for nanoMIPS CPUs (5/5)Philippe Mathieu-Daudé
Part 5/5: Convert jumping to kernel Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-11-philmd@linaro.org>
2023-01-13hw/mips/malta: Use bootloader generator API for nanoMIPS CPUs (4/5)Philippe Mathieu-Daudé
Part 4/5: Convert GT64120 ISD base address setup Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-10-philmd@linaro.org>
2023-01-13hw/mips/malta: Use bootloader generator API for nanoMIPS CPUs (3/5)Philippe Mathieu-Daudé
Part 3/5: Convert PCI0 I/O BAR setup Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-9-philmd@linaro.org>
2023-01-13hw/mips/malta: Use bootloader generator API for nanoMIPS CPUs (2/5)Philippe Mathieu-Daudé
Part 2/5: Convert PCI0 MEM0 BAR setup Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-8-philmd@linaro.org>
2023-01-13hw/mips/malta: Use bootloader generator API for nanoMIPS CPUs (1/5)Philippe Mathieu-Daudé
Similarly to how commit 0c8427baf0 ("hw/mips/malta: Use bootloader helper to set BAR registers") converted write_bootloader(), convert the equivalent write_bootloader_nanomips(), allowing us to modify the bootloader code more easily in the future. Part 1/5: Convert PCI0 MEM1 BAR setup Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-7-philmd@linaro.org>
2023-01-13hw/mips/bootloader: Implement nanoMIPS JALRc opcode generatorPhilippe Mathieu-Daudé
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-6-philmd@linaro.org>
2023-01-13hw/mips/bootloader: Implement nanoMIPS LI (LUI+ORI) opcode generatorPhilippe Mathieu-Daudé
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-5-philmd@linaro.org>
2023-01-13hw/mips/bootloader: Implement nanoMIPS SW opcode generatorPhilippe Mathieu-Daudé
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-4-philmd@linaro.org>
2023-01-13hw/mips/bootloader: Implement nanoMIPS NOP opcode generatorPhilippe Mathieu-Daudé
Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <20221211204533.85359-3-philmd@linaro.org>
2023-01-13hw/mips/bootloader: Handle buffers as opaque arraysPhilippe Mathieu-Daudé
It is irrelevant to the API what the buffers to fill are made of. In particular, some MIPS ISA have 16-bit wide instructions. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20221211204533.85359-2-philmd@linaro.org>
2023-01-13tests/avocado: Add tests booting YAMON ROM on MIPS Malta machinesPhilippe Mathieu-Daudé
Add quick tests booting YAMON: $ avocado --show=app,console run -t machine:malta tests/avocado/machine_mips_malta.py (1/2) tests/avocado/machine_mips_malta.py:MaltaMachine.test_mipsel_malta_yamon: console: YAMON ROM Monitor, Revision 02.22. console: Copyright (c) 1999-2007 MIPS Technologies, Inc. - All Rights Reserved. console: For a list of available commands, type 'help'. console: Compilation time = May 24 2013 12:16:34 (pburton) console: Board type/revision = 0x02 (Malta) / 0x00 console: Core board type/revision = 0x01 (CoreLV) / 0x00 console: System controller/revision = Galileo / GT_64120A-B-0 console: FPGA revision = 0x0000 console: MAC address = ff.ff.ff.ff.ff.ff console: Board S/N = 0123456789 console: PCI bus frequency = 33.33 MHz console: Processor Company ID/options = 0x01 (MIPS Technologies, Inc.) / 0x00 console: Processor ID/revision = 0x93 (MIPS 24Kf) / 0x00 console: Endianness = Little console: CPU/Bus frequency = 333 MHz / 419 MHz console: Coherency = None console: Flash memory size = 4 MByte console: SDRAM size = 128 MByte console: First free SDRAM address = 0x800c32f0 console: WARNING: Environment variable flash area is invalid! console: HINT : Perform "erase -e" console: YAMON> PASS (1.88 s) (2/2) tests/avocado/machine_mips_malta.py:MaltaMachine.test_mips64el_malta_yamon: ... console: System controller/revision = Galileo / GT_64120A-B-0 console: Processor Company ID/options = 0x01 (MIPS Technologies, Inc.) / 0x00 console: Processor ID/revision = 0x82 (MIPS 20Kc) / 0xa0 ... console: YAMON> PASS (1.89 s) RESULTS : PASS 2 | ERROR 0 | FAIL 0 | SKIP 0 | WARN 0 | INTERRUPT 0 | CANCEL 0 JOB TIME : 4.57 s YAMON does some endian-swapped acceses on the ISD<->PCI CFG/DATA registers. These tests are useful to debug cross-endianness issues, in particular on big-endian host. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230104133935.4639-7-philmd@linaro.org>
2023-01-13hw/mips/gt64xxx_pci: Move it to hw/pci-host/Philippe Mathieu-Daudé
The GT-64120 is a north-bridge, and it is not MIPS specific. Move it with the other north-bridge devices. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-Id: <20221209151533.69516-8-philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2023-01-13hw/mips/meson: Make gt64xxx_pci.c endian-agnosticPhilippe Mathieu-Daudé
The single machine using this device explicitly sets its endianness. We don't need to set a default. This allow us to remove the target specificity from the build system. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-Id: <20221209151533.69516-7-philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2023-01-13hw/mips/malta: Explicit GT64120 endianness upon device creationPhilippe Mathieu-Daudé
Propagate the controller endianess from the machine, setting the "cpu-little-endian" property. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <20221209151533.69516-6-philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2023-01-13hw/mips/gt64xxx_pci: Add a 'cpu-little-endian' qdev propertyPhilippe Mathieu-Daudé
This device does not have to be TARGET-dependent. Add a 'cpu_big_endian' property which sets the byte-swapping options if required. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-Id: <20221220113436.14299-5-philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2023-01-13hw/mips/gt64xxx_pci: Manage endian bits with the RegisterFields APIPhilippe Mathieu-Daudé
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <20221220113436.14299-4-philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2023-01-13hw/mips/gt64xxx_pci: Let the GT64120 manage the lower 512MiB holePhilippe Mathieu-Daudé
Per the comment in the Malta board, the [0x0000.0000-0x2000.0000] range is decoded by the GT64120, so move the "empty_slot" there. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-Id: <20221209151533.69516-3-philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2023-01-13hw/mips/Kconfig: Introduce CONFIG_GT64120 to select gt64xxx_pci.cPhilippe Mathieu-Daudé
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Bernhard Beschow <shentey@gmail.com> Message-Id: <20221209151533.69516-2-philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2023-01-13hw/mips/gt64xxx_pci: Endian-swap using PCI_HOST_BRIDGE MemoryRegionOpsPhilippe Mathieu-Daudé
GT64120's PCI endianness swapping works on little-endian hosts, but doesn't on big-endian ones. Instead of complicating how CFGADDR/CFGDATA registers deal with endianness, use the existing MemoryRegionOps from hw/pci/pci_host.c. Doing so also reduce the access to internal PCI_HOST_BRIDGE fields. Map the PCI_HOST_BRIDGE MemoryRegionOps into the corresponding CFGADDR/CFGDATA regions in the ISD MMIO and remove the unused code in the current ISD read/write handlers. Update the mapping when PCI0_CMD register is accessed (in case the endianness is changed). This allows using the GT64120 on a big-endian host (and boot the MIPS Malta machine in little-endian). Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230104133935.4639-6-philmd@linaro.org>
2023-01-13hw/mips/gt64xxx_pci: Accumulate address space changesPhilippe Mathieu-Daudé
Single registers access in ISD can produce multiple changes in the address spaces. To reduce computational effort, accumulate these as a single memory transaction. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230104133935.4639-5-philmd@linaro.org>
2023-01-13hw/mips/malta: Trace FPGA LEDs/ASCII display updatesPhilippe Mathieu-Daudé
The FPGA LEDs/ASCII display is mostly used by the bootloader to show very low-level debug info. QEMU connects its output to a character device backend, which is not very practical to correlate with ASM instruction executed, interrupts or MMIO accesses. Also, the display discard the previous states. To ease bootloader debugging experience, add a pair of trace events. Such events can be analyzed over time or diff-ed between different runs. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230104133935.4639-4-philmd@linaro.org>
2023-01-13hw/mips/malta: Split FPGA LEDs/ASCII display updatesPhilippe Mathieu-Daudé
No need to refresh the ASCII bar when a LED is toggled (and vice versa). Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230104133935.4639-3-philmd@linaro.org>
2023-01-12Merge tag 'for-upstream' of https://gitlab.com/bonzini/qemu into stagingPeter Maydell
* Atomic memslot updates for KVM (Emanuele, David) * libvhost-user/libvduse warnings fixes (Marcel) * i386 TCG fixes (Joe, myself) * Remove compilation errors when -Werror=maybe-uninitialized (Eric) * fix GLIB_VERSION for cross-compilation (Paolo) # -----BEGIN PGP SIGNATURE----- # # iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAmO+hUYUHHBib256aW5p # QHJlZGhhdC5jb20ACgkQv/vSX3jHroM5+Qf9HIy2QDdGZJ1fA3AUhp7bCJgotlRF # 2NMFwo3P/Vwe4qMzzRGkMpsDLFhJx6j1V+sSyeLNxCPtidHccFcmKJOlTqmGeT0k # HccK07cgAXSZUtvog2xid7PeNpi8OQjKih2Gn/FnN9Tjtad7yMYfXO3O5ttkRV+7 # F8qXBkm4J74l5zs+UMOXUj9y4PudjRFPS0q/lti82eg8mUZ33BZtBkP18Qjq45GM # wLcemfId0Z6e4s4fmVeTobZ31ltz/B6HDkoD59pQSK6QW2VBIAoO64Hw6FdZnu5W # PltBsDy5cboKHibkb/ySZQUMhb7UemKFyf7eRWkdlnSSHHFdL/K4x7x7DA== # =0b+g # -----END PGP SIGNATURE----- # gpg: Signature made Wed 11 Jan 2023 09:45:42 GMT # gpg: using RSA key F13338574B662389866C7682BFFBD25F78C7AE83 # gpg: issuer "pbonzini@redhat.com" # gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>" [full] # gpg: aka "Paolo Bonzini <pbonzini@redhat.com>" [full] # Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4 E2F7 7E15 100C CD36 69B1 # Subkey fingerprint: F133 3857 4B66 2389 866C 7682 BFFB D25F 78C7 AE83 * tag 'for-upstream' of https://gitlab.com/bonzini/qemu: (29 commits) configure: remove backwards-compatibility code target/i386: fix operand size of unary SSE operations libvduse: Add extra compiler warnings libvhost-user: Add extra compiler warnings libvhost-user: Fix assignment in vring_set_avail_event libvduse: Fix assignment in vring_set_avail_event libvduse: Switch to unsigned int for inuse field in struct VduseVirtq libvduse: Provide _GNU_SOURCE when compiling outside of QEMU libvhost-user: Change dev->postcopy_ufd assignment to make it C90 compliant libvhost-user: Declare uffdio_register early to make it C90 compliant libvhost-user: Use unsigned int i for some for-loop iterations libvhost-user: Cast rc variable to avoid compiler warning libvhost-user: Replace typeof with __typeof__ libvhost-user: Provide _GNU_SOURCE when compiling outside of QEMU hw/display: avoid creating empty loadable modules enforce use of G_GNUC_PRINTF attributes tests: add G_GNUC_PRINTF for various functions util/error: add G_GNUC_PRINTF for various functions tools/virtiofsd: add G_GNUC_PRINTF for logging functions hw/xen: use G_GNUC_PRINTF/SCANF for various functions ... Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2023-01-11Merge tag 'nvme-next-pull-request' of https://gitlab.com/birkelund/qemu into ↵Peter Maydell
staging hw/nvme updates # -----BEGIN PGP SIGNATURE----- # # iQEzBAABCgAdFiEEUigzqnXi3OaiR2bATeGvMW1PDekFAmO+anUACgkQTeGvMW1P # Del4HAf/eqW92bf04CpXPzLMG38+FoYQgwkH8Lp390JlSM6kjv3DgPM2ygJXIEYr # dV0KQpTcPRGAVCuXCSOUd6FTkAYdcPVtOeJeYhldEYj9cc9Y/9tvcBMthFBkiCU/ # 66TL7JEgFmOfVMrLypYpfrPHCN0lQVX+i2+Qruospg6itubAKcYdJq+Hx2WWHzcC # hl/qNgSSIGlhQi21ada1iCUC8Q3TZzo54M+0tW4zfybzsG+eNM+ouKDl/qOHtQgP # ON/ef9dl+0Dum1l5z7O+sPiXBzZRJR7V+jZOsbf0A98fxqmuu3YacVrH6mY/Zmdx # e1xexJt6PVXaIKtoZQKRVn2C0ITjXA== # =AvaO # -----END PGP SIGNATURE----- # gpg: Signature made Wed 11 Jan 2023 07:51:17 GMT # gpg: using RSA key 522833AA75E2DCE6A24766C04DE1AF316D4F0DE9 # gpg: Good signature from "Klaus Jensen <its@irrelevant.dk>" [full] # gpg: aka "Klaus Jensen <k.jensen@samsung.com>" [full] # Primary key fingerprint: DDCA 4D9C 9EF9 31CC 3468 4272 63D5 6FC5 E55D A838 # Subkey fingerprint: 5228 33AA 75E2 DCE6 A247 66C0 4DE1 AF31 6D4F 0DE9 * tag 'nvme-next-pull-request' of https://gitlab.com/birkelund/qemu: hw/nvme: cleanup error reporting in nvme_init_pci() hw/nvme: clean up confusing use of errp/local_err hw/nvme: fix missing cq eventidx update hw/nvme: fix missing endian conversions for doorbell buffers hw/nvme: rename shadow doorbell related trace events hw/nvme: use QOM accessors Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2023-01-11configure: remove backwards-compatibility codePaolo Bonzini
The cmd_line.txt mangling is only needed when rebuilding from very old trees and is kept mostly as an example of how to extend it. However, Meson 0.63 introduces a deprecation mechanism for meson_options.txt that can be used instead, so get rid of our home-grown hack. Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11target/i386: fix operand size of unary SSE operationsPaolo Bonzini
VRCPSS, VRSQRTSS and VCVTSx2Sx have a 32-bit or 64-bit memory operand, which is represented in the decoding tables by X86_VEX_REPScalar. Add it to the tables, and make validate_vex() handle the case of an instruction that is in exception type 4 without the REP prefix and exception type 5 with it; this is the cas of VRCP and VRSQRT. Reported-by: yongwoo <https://gitlab.com/yongwoo36> Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1377 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvduse: Add extra compiler warningsMarcel Holtmann
In case libvhost-user is used externally, that projects compiler warnings might be more strict. Enforce an extra set of compiler warnings to catch issues early on. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Message-Id: <08daa1896ad8824e17d57d6a970bc0b4bee73ece.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Add extra compiler warningsMarcel Holtmann
In case libvhost-user is used externally, that projects compiler warnings might be more strict. Enforce an extra set of compiler warnings to catch issues early on. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Message-Id: <737ebf2e697f8640558e6f73d96a692711f548f6.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Fix assignment in vring_set_avail_eventMarcel Holtmann
Since it was proposed to change the code in libvduse.c to use memcpy instead of an assignment, the code in libvhost-user.c should also be changed to use memcpy. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Message-Id: <502b22723264db064e4b05008233a9c1f2f8aaaa.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvduse: Fix assignment in vring_set_avail_eventMarcel Holtmann
Since the assignment is causing a compiler warning, fix it by using memcpy instead. CC libvduse.o libvduse.c: In function ‘vring_set_avail_event’: libvduse.c:603:7: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasin] 603 | *((uint16_t *)&vq->vring.used->ring[vq->vring.num]) = htole16(val); | ~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Suggested-by: Xie Yongji <xieyongji@bytedance.com> Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Message-Id: <4a0fe2a6436464473119fdbf0bc4076b36fbb37f.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvduse: Switch to unsigned int for inuse field in struct VduseVirtqMarcel Holtmann
It seems there is no need to keep the inuse field signed and end up with compiler warnings for sign-compare. CC libvduse.o libvduse.c: In function ‘vduse_queue_pop’: libvduse.c:789:19: error: comparison of integer expressions of different signedness: ‘int’ and ‘unsigned int’ [-Werror=sign-compare] 789 | if (vq->inuse >= vq->vring.num) { | ^~ Instead of casting the comparison to unsigned int, just make the inuse field unsigned int in the fist place. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Reviewed-by: Xie Yongji <xieyongji@bytedance.com> Message-Id: <9fe3fd8b042e048bd04d506ca6e43d738b5c45b7.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvduse: Provide _GNU_SOURCE when compiling outside of QEMUMarcel Holtmann
When the libvduse sources are used by another project, it can not be guaranteed that _GNU_SOURCE is set by the build system. If it is for example not set, errors like this show up. CC libvduse.o libvduse.c: In function ‘vduse_log_get’: libvduse.c:172:9: error: implicit declaration of function ‘ftruncate’; did you mean ‘strncat’? [-Werror=implicit-function-declaration] 172 | if (ftruncate(fd, size) == -1) { | ^~~~~~~~~ | strncat The simplest way to allow external complication of libvduse.[ch] by setting _GNU_SOURCE if it is not already set by the build system. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Message-Id: <407f3665f0605df936e5bfe60831d180edfb8cca.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Change dev->postcopy_ufd assignment to make it C90 compliantMarcel Holtmann
The assignment of dev->postcopy_ufd can be moved into an else clause and then the code becomes C90 compliant. CC libvhost-user.o libvhost-user.c: In function ‘vu_set_postcopy_advise’: libvhost-user.c:1625:5: error: ISO C90 forbids mixed declarations and code [-Werror=declaration-after-statement] 1625 | struct uffdio_api api_struct; | ^~~~~~ Understandable, it might be desired to avoid else clauses, but in this case it seems clear enough and frankly the dev->postcopy_ufd is only assigned once. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <74db52afb1203c4580ffc7fa462b4b2ba260a353.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Declare uffdio_register early to make it C90 compliantMarcel Holtmann
When using libvhost-user source in an external project that wants to comply with the C90 standard, it is best to declare variables before code. CC libvhost-user.o libvhost-user.c: In function ‘generate_faults’: libvhost-user.c:683:9: error: ISO C90 forbids mixed declarations and code [-Werror=declaration-after-statement] 683 | struct uffdio_register reg_struct; | ^~~~~~ In this case, it is also simple enough and doesn't cause any extra ifdef additions. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <556c2d00c01fa134d13c0371d4014c90694c2943.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Use unsigned int i for some for-loop iterationsMarcel Holtmann
The sign-compare warning also hits some of the for-loops, but it easy fixed by just making the iterator variable unsigned int. CC libvhost-user.o libvhost-user.c: In function ‘vu_gpa_to_va’: libvhost-user.c:223:19: error: comparison of integer expressions of different signedness: ‘int’ and ‘uint32_t’ {aka ‘unsigned int’} [-Werror=sign-compare] 223 | for (i = 0; i < dev->nregions; i++) { | ^ Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Message-Id: <decb925e1a6fb9538738d2570bda2804f888fa15.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Cast rc variable to avoid compiler warningMarcel Holtmann
The assert from recvmsg() return value against an uint32_t size field from a protocol struct throws a compiler warning. CC libvhost-user.o In file included from libvhost-user.c:27: libvhost-user.c: In function ‘vu_message_read_default’: libvhost-user.c:363:19: error: comparison of integer expressions of different signedness: ‘int’ and ‘uint32_t’ {aka ‘unsigned int’} [-Werror=sign-compare] 363 | assert(rc == vmsg->size); | ^~ This is not critical, but annoying when the libvhost-user source are used in an external project that has this compiler warning switched on. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Message-Id: <7a791e27b7bd3e0a8b8cc8fbb15090a870d226d5.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Replace typeof with __typeof__Marcel Holtmann
Strictly speaking only -std=gnu99 support the usage of typeof and for easier inclusion in external projects, it is better to use __typeof__. CC libvhost-user.o libvhost-user.c: In function ‘vu_log_queue_fill’: libvhost-user.c:86:13: error: implicit declaration of function ‘typeof’ [-Werror=implicit-function-declaration] 86 | typeof(x) _min1 = (x); \ | ^~~~~~ Changing these two users of typeof makes the compiler happy and no extra flags or pragmas need to be provided. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <981aa822bcaaa2b8d74f245339a99a85c25b346f.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11libvhost-user: Provide _GNU_SOURCE when compiling outside of QEMUMarcel Holtmann
Then the libvhost-user sources are used by another project, it can not be guaranteed that _GNU_SOURCE is set by the build system. If it is for example not set, errors like this show up. CC libvhost-user.o libvhost-user.c: In function ‘vu_panic’: libvhost-user.c:195:9: error: implicit declaration of function ‘vasprintf’; did you mean ‘vsprintf’? [-Werror=implicit-function-declaration] 195 | if (vasprintf(&buf, msg, ap) < 0) { | ^~~~~~~~~ | vsprintf The simplest way to allow external complication of libvhost-user.[ch] is by setting _GNU_SOURCE if it is not already set by the build system. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Message-Id: <be27dcc747a6b5cc6f8ae3f79e0b79171382bcef.1671741278.git.marcel@holtmann.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11hw/display: avoid creating empty loadable modulesDaniel P. Berrangé
When using --disable-virglrenderer, QEMU still creates hw-display-virtio-gpu-gl.so hw-display-virtio-vga-gl.so hw-display-virtio-gpu-pci-gl.so but when these are loaded, they provide no functionality as the code which registers types is not compiled in. Funtionally this is relatively harmless, because QEMU is fine loading a module with no types. This is rather confusing for users and OS distro maintainers though, as they think they have the GL functionality built, but in fact the module they are looking at provides nothing of value. The root cause is the use of 'when/if_true' rules when adding sources to the module source set. If all the rules evaluate to false, then we have declared the module, but not added anything to it. We need to put declaration of the entire module inside a condition based on existance of the 3rd party library deps that are mandatory. Fixes: https://gitlab.com/qemu-project/qemu/-/issues/1352 Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Message-Id: <20221219125830.2369169-1-berrange@redhat.com> [Do not check for pixman. - Paolo] Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11enforce use of G_GNUC_PRINTF attributesDaniel P. Berrangé
We've been very gradually adding G_GNUC_PRINTF annotations to functions over years. This has been useful in detecting certain malformed printf strings, or cases where we pass user data as the printf format which is a potential security flaw. Given the inherant memory corruption danger in use of format strings vs mis-matched variadic arguments, it is worth applying G_GNUC_PRINTF to all functions using printf, even if we know they are safe. The compilers can reasonably reliably identify such places with the -Wsuggest-attribute=format / -Wmissing-format-attribute flags. Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Message-Id: <20221219130205.687815-7-berrange@redhat.com> [-Wsuggest-attribute=format and -Wmissing-format-attribute are synonyms, only include one; disable it for testfloat. - Paolo] Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11tests: add G_GNUC_PRINTF for various functionsDaniel P. Berrangé
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Message-Id: <20221219130205.687815-6-berrange@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11util/error: add G_GNUC_PRINTF for various functionsDaniel P. Berrangé
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <20221219130205.687815-5-berrange@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11tools/virtiofsd: add G_GNUC_PRINTF for logging functionsDaniel P. Berrangé
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Message-Id: <20221219130205.687815-4-berrange@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11hw/xen: use G_GNUC_PRINTF/SCANF for various functionsDaniel P. Berrangé
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Acked-by: Anthony PERARD <anthony.perard@citrix.com> Message-Id: <20221219130205.687815-3-berrange@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11disas: add G_GNUC_PRINTF to gstring_printfDaniel P. Berrangé
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Message-Id: <20221219130205.687815-2-berrange@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11gitlab: remove redundant setting of PKG_CONFIG_PATHDaniel P. Berrangé
The PKG_CONFIG_PATH variable is not defined in GitLab CI envs and even if it was, we don't need to set it to its existing value. Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Message-Id: <20221103173044.3969425-2-berrange@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11chardev: clean up chardev-parallel.cPaolo Bonzini
Replace HAVE_CHARDEV_PARPORT with a Meson conditional, remove unnecessary defines, and close the file descriptor on FreeBSD/DragonFly. Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11target/i386: Remove compilation errors when -Werror=maybe-uninitializedEric Auger
To avoid compilation errors when -Werror=maybe-uninitialized is used, add a default case with g_assert_not_reached(). Otherwise with GCC 11.3.1 "cc (GCC) 11.3.1 20220421 (Red Hat 11.3.1-2)" we get: ../target/i386/ops_sse.h: In function ‘helper_vpermdq_ymm’: ../target/i386/ops_sse.h:2495:13: error: ‘r3’ may be used uninitialized in this function [-Werror=maybe-uninitialized] 2495 | d->Q(3) = r3; | ~~~~~~~~^~~~ ../target/i386/ops_sse.h:2494:13: error: ‘r2’ may be used uninitialized in this function [-Werror=maybe-uninitialized] 2494 | d->Q(2) = r2; | ~~~~~~~~^~~~ ../target/i386/ops_sse.h:2493:13: error: ‘r1’ may be used uninitialized in this function [-Werror=maybe-uninitialized] 2493 | d->Q(1) = r1; | ~~~~~~~~^~~~ ../target/i386/ops_sse.h:2492:13: error: ‘r0’ may be used uninitialized in this function [-Werror=maybe-uninitialized] 2492 | d->Q(0) = r0; | ~~~~~~~~^~~~ Signed-off-by: Eric Auger <eric.auger@redhat.com> Message-Id: <20221222140158.1260748-1-eric.auger@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2023-01-11kvm: Atomic memslot updatesDavid Hildenbrand
If we update an existing memslot (e.g., resize, split), we temporarily remove the memslot to re-add it immediately afterwards. These updates are not atomic, especially not for KVM VCPU threads, such that we can get spurious faults. Let's inhibit most KVM ioctls while performing relevant updates, such that we can perform the update just as if it would happen atomically without additional kernel support. We capture the add/del changes and apply them in the notifier commit stage instead. There, we can check for overlaps and perform the ioctl inhibiting only if really required (-> overlap). To keep things simple we don't perform additional checks that wouldn't actually result in an overlap -- such as !RAM memory regions in some cases (see kvm_set_phys_mem()). To minimize cache-line bouncing, use a separate indicator (in_ioctl_lock) per CPU. Also, make sure to hold the kvm_slots_lock while performing both actions (removing+re-adding). We have to wait until all IOCTLs were exited and block new ones from getting executed. This approach cannot result in a deadlock as long as the inhibitor does not hold any locks that might hinder an IOCTL from getting finished and exited - something fairly unusual. The inhibitor will always hold the BQL. AFAIKs, one possible candidate would be userfaultfd. If a page cannot be placed (e.g., during postcopy), because we're waiting for a lock, or if the userfaultfd thread cannot process a fault, because it is waiting for a lock, there could be a deadlock. However, the BQL is not applicable here, because any other guest memory access while holding the BQL would already result in a deadlock. Nothing else in the kernel should block forever and wait for userspace intervention. Note: pause_all_vcpus()/resume_all_vcpus() or start_exclusive()/end_exclusive() cannot be used, as they either drop the BQL or require to be called without the BQL - something inhibitors cannot handle. We need a low-level locking mechanism that is deadlock-free even when not releasing the BQL. Signed-off-by: David Hildenbrand <david@redhat.com> Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com> Tested-by: Emanuele Giuseppe Esposito <eesposit@redhat.com> Message-Id: <20221111154758.1372674-4-eesposit@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
generated by cgit v1.2.3 (git 2.26.2) at 2025-02-23 10:58:32 +0000