host: Support macOS 12
Subsystems (2)
v7 cover host: Support macOS 12
From: Philippe Mathieu-Daudé <[email protected]>
Few patches to be able to build QEMU on macOS 12 (Monterey).
Missing review:
0006-hvf-Fix-OOB-write-in-RDTSCP-instruction-decode.patch 0013-osdep-Avoid-using-Clang-specific-__builtin_available.patch 0014-meson-Resolve-the-entitlement.sh-script-once-for-goo.patch 0015-meson-Log-QEMU_CXXFLAGS-content-in-summary.patch 0016-configure-Pass-filtered-QEMU_OBJCFLAGS-to-meson.patch 0017-ui-cocoa-Constify-qkeycode-translation-arrays.patch 0020-ui-cocoa-capture-all-keys-and-combos-when-mouse-is-g.patch 0021-ui-cocoa-add-option-to-swap-Option-and-Command.patch 0022-gitlab-ci-Support-macOS-12-via-cirrus-run.patch
Since v6:
- Dropped merged patches
- Addressed Akihiko Odaki comments (squashed 2 patches, added R/T-b)
- Dropped 'configure: Disable out-of-line atomic operations on Aarch64'
- Add few macos patches on the list pending for 7.0 so tested by CI
Since v5:
- Fixed failed rebase between patches 10 and 16 (Akihiko)
- Include "ui/cocoa: Fix the leak of qemu_console_get_label"
Since v4:
- Use MAC_OS_X_VERSION_MIN_REQUIRED definition (Akihiko)
- Include patches from Akihiko
Since v3:
- Fix --enable-modules
- Ignore #pragma on softfloat3 tests
- Addressed Akihiko Odaki comments
- Include Cameron Esfahani patches
Since v2:
- Addressed Akihiko Odaki comments:
. use __is_identifier(),
. remove cocoa setAllowedFileTypes()
- Addressed Daniel Berrangé comment:
. rebased on testing/next, update libvirt-ci/lcitool
Akihiko Odaki (2):
audio: Log context for audio bug
coreaudio: Always return 0 in handle_voice_change
Cameron Esfahani (2):
hvf: Use standard CR0 and CR4 register definitions
hvf: Fix OOB write in RDTSCP instruction decode
Carwyn Ellis (2):
ui/cocoa: add option to disable left-command forwarding to guest ui/cocoa: release mouse when user switches away from QEMU window
Gustavo Noronha Silva (2):
ui/cocoa: capture all keys and combos when mouse is grabbed ui/cocoa: add option to swap Option and Command
Philippe Mathieu-Daudé (14):
configure: Allow passing extra Objective C compiler flags tests/fp/berkeley-testfloat-3: Ignore ignored #pragma directives
hvf: Make hvf_get_segments() / hvf_put_segments() local
hvf: Remove deprecated hv_vcpu_flush() calls block/file-posix: Remove a deprecation warning on macOS 12 audio/coreaudio: Remove a deprecation warning on macOS 12 audio/dbus: Fix building with modules on macOS
audio: Rename coreaudio extension to use Objective-C compiler
osdep: Avoid using Clang-specific __builtin_available()
meson: Resolve the entitlement.sh script once for good
meson: Log QEMU_CXXFLAGS content in summary
configure: Pass filtered QEMU_OBJCFLAGS to meson ui/cocoa: Constify qkeycode translation arrays
gitlab-ci: Support macOS 12 via cirrus-run
.gitlab-ci.d/cirrus.yml | 16 ++++
.gitlab-ci.d/cirrus/macos-12.vars | 16 ++++
audio/audio.c | 25 +++---
audio/audio_template.h | 27 +++----
audio/{coreaudio.c => coreaudio.m} | 23 +++---
audio/meson.build | 4 +-
block/file-posix.c | 14 +++-
configure | 31 ++++++++
include/qemu/osdep.h | 10 +--
meson.build | 17 +++-
qapi/ui.json | 29 +++++++
qemu-options.hx | 15 ++++
target/i386/hvf/vmx.h | 19 +++--
target/i386/hvf/x86.c | 6 +-
target/i386/hvf/x86.h | 34 --------
target/i386/hvf/x86_decode.c | 12 ++-
target/i386/hvf/x86_mmu.c | 2 +-
target/i386/hvf/x86_task.c | 4 +-
target/i386/hvf/x86hvf.c | 14 +++-
target/i386/hvf/x86hvf.h | 3 +-
tests/fp/meson.build | 5 ++
tests/lcitool/refresh | 1 +
ui/cocoa.m | 122 ++++++++++++++++++++++++++---
23 files changed, 327 insertions(+), 122 deletions(-)
create mode 100644 .gitlab-ci.d/cirrus/macos-12.vars
rename audio/{coreaudio.c => coreaudio.m} (97%)Signature · 1 line
2.34.1
PATCH v7 1/22 configure: Allow passing extra Objective C compiler flags
From: Philippe Mathieu-Daudé <[email protected]>
We can pass C/CPP/LD flags via CFLAGS/CXXFLAGS/LDFLAGS environment variables, or via configure --extra-cflags / --extra-cxxflags / --extra-ldflags options. Provide similar behavior for Objective C:
use existing flags from $OBJCFLAGS, or passed via --extra-objcflags.
--- configure | 8 ++++++++ meson.build | 5 +++++ 2 files changed, 13 insertions(+)
configure +8 -0
@@ -285,6 +285,7 @@ done EXTRA_CFLAGS="" EXTRA_CXXFLAGS="" +EXTRA_OBJCFLAGS="" EXTRA_LDFLAGS="" xen_ctrl_version="$default_feature"
@@ -366,9 +367,12 @@ for opt do --extra-cflags=*) EXTRA_CFLAGS="$EXTRA_CFLAGS $optarg" EXTRA_CXXFLAGS="$EXTRA_CXXFLAGS $optarg" + EXTRA_OBJCFLAGS="$EXTRA_OBJCFLAGS $optarg" ;; --extra-cxxflags=*) EXTRA_CXXFLAGS="$EXTRA_CXXFLAGS $optarg" ;; + --extra-objcflags=*) EXTRA_OBJCFLAGS="$EXTRA_OBJCFLAGS $optarg" + ;; --extra-ldflags=*) EXTRA_LDFLAGS="$EXTRA_LDFLAGS $optarg" ;; --enable-debug-info) debug_info="yes"
@@ -748,6 +752,8 @@ for opt do ;; --extra-cxxflags=*) ;; + --extra-objcflags=*) + ;; --extra-ldflags=*) ;; --enable-debug-info)
@@ -1172,6 +1178,7 @@ Advanced options (experts only): --objcc=OBJCC use Objective-C compiler OBJCC [$objcc] --extra-cflags=CFLAGS append extra C compiler flags CFLAGS --extra-cxxflags=CXXFLAGS append extra C++ compiler flags CXXFLAGS + --extra-objcflags=OBJCFLAGS append extra Objective C compiler flags OBJCFLAGS --extra-ldflags=LDFLAGS append extra linker flags LDFLAGS --cross-cc-ARCH=CC use compiler when building ARCH guest test cases --cross-cc-cflags-ARCH= use compiler flags when building ARCH guest tests
@@ -3137,6 +3144,7 @@ if test "$skip_meson" = no; then echo "[built-in options]" >> $cross echo "c_args = [$(meson_quote $CFLAGS $EXTRA_CFLAGS)]" >> $cross echo "cpp_args = [$(meson_quote $CXXFLAGS $EXTRA_CXXFLAGS)]" >> $cross + test -n "$objcc" && echo "objc_args = [$(meson_quote $OBJCFLAGS $EXTRA_OBJCFLAGS)]" >> $cross echo "c_link_args = [$(meson_quote $CFLAGS $LDFLAGS $EXTRA_CFLAGS $EXTRA_LDFLAGS)]" >> $cross echo "cpp_link_args = [$(meson_quote $CXXFLAGS $LDFLAGS $EXTRA_CXXFLAGS $EXTRA_LDFLAGS)]" >> $cross echo "[binaries]" >> $cross
meson.build +5 -0
@@ -3457,6 +3457,11 @@ if link_language == 'cpp' + ['-O' + get_option('optimization')] + (get_option('debug') ? ['-g'] : []))} endif +if targetos == 'darwin' + summary_info += {'OBJCFLAGS': ' '.join(get_option('objc_args') + + ['-O' + get_option('optimization')] + + (get_option('debug') ? ['-g'] : []))} +endif link_args = get_option(link_language + '_link_args') if link_args.length() > 0 summary_info += {'LDFLAGS': ' '.join(link_args)}
Signature · 1 line
2.34.1
PATCH v7 2/22 tests/fp/berkeley-testfloat-3: Ignore ignored #pragma directives
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Since we already use -Wno-unknown-pragmas, we can also use -Wno-ignored-pragmas. This silences hundred of warnings using clang 13 on macOS Monterey:
[409/771] Compiling C object tests/fp/libtestfloat.a.p/berkeley-testfloat-3_source_test_az_f128_rx.c.o ../tests/fp/berkeley-testfloat-3/source/test_az_f128_rx.c:49:14: warning: '#pragma FENV_ACCESS' is not supported on this target - ignored [-Wignored-pragmas] #pragma STDC FENV_ACCESS ON
^
1 warning generated.
Having:
$ cc -v Apple clang version 13.0.0 (clang-1300.0.29.30)
--- tests/fp/meson.build | 5 +++++ 1 file changed, 5 insertions(+)
tests/fp/meson.build +5 -0
@@ -37,6 +37,11 @@ tfcflags = [ '-Wno-error', ] +if cc.get_id() == 'clang' + # Clang does not support '#pragma STDC FENV_ACCESS' + tfcflags += [ '-Wno-ignored-pragmas' ] +endif + tfgencases = [ tfdir / 'genCases_ui32.c', tfdir / 'genCases_ui64.c',
Signature · 1 line
2.34.1
PATCH v7 3/22 hvf: Use standard CR0 and CR4 register definitions
In reply to Philippe Mathieu-DaudéFrom: Cameron Esfahani <[email protected]>
No need to have our own definitions of these registers.
--- target/i386/hvf/vmx.h | 17 +++++++++-------- target/i386/hvf/x86.c | 6 +++--- target/i386/hvf/x86.h | 34 ---------------------------------- target/i386/hvf/x86_mmu.c | 2 +- target/i386/hvf/x86_task.c | 3 ++- 5 files changed, 15 insertions(+), 47 deletions(-)
target/i386/hvf/vmx.h +9 -8
@@ -124,10 +124,11 @@ static inline void macvm_set_cr0(hv_vcpuid_t vcpu, uint64_t cr0)uint64_t efer =rvmcs (vcpu,VMCS_GUEST_IA32_EFER );uint64_t old_cr0 =rvmcs (vcpu,VMCS_GUEST_CR0 );uint64_t changed_cr0 = old_cr0 ^ cr0; -uint64_t mask =CR0_PG |CR0_CD |CR0_NW |CR0_NE |CR0_ET ; +uint64_t mask =CR0_PG_MASK |CR0_CD_MASK |CR0_NW_MASK | +CR0_NE_MASK |CR0_ET_MASK ;uint64_t entry_ctls; -if ((cr0 &CR0_PG ) && (rvmcs (vcpu,VMCS_GUEST_CR4 ) &CR4_PAE ) && +if ((cr0 &CR0_PG_MASK ) && (rvmcs (vcpu,VMCS_GUEST_CR4 ) &CR4_PAE_MASK ) && !(efer &MSR_EFER_LME )) { address_space_read(&address_space_memory, rvmcs(vcpu, VMCS_GUEST_CR3) & ~0x1f,
@@ -142,8 +143,8 @@ static inline void macvm_set_cr0(hv_vcpuid_t vcpu, uint64_t cr0)wvmcs (vcpu,VMCS_CR0_SHADOW , cr0);if (efer &MSR_EFER_LME ) { -if (changed_cr0 &CR0_PG ) { -if (cr0 &CR0_PG ) { +if (changed_cr0 &CR0_PG_MASK ) { +if (cr0 &CR0_PG_MASK ) {enter_long_mode (vcpu, cr0, efer); }else {exit_long_mode (vcpu, cr0, efer);
@@ -155,8 +156,8 @@ static inline void macvm_set_cr0(hv_vcpuid_t vcpu, uint64_t cr0) }/* Filter new CR0 after we are finished examining it above. */ - cr0 = (cr0 & ~(mask & ~CR0_PG )); -wvmcs (vcpu,VMCS_GUEST_CR0 , cr0 |CR0_NE |CR0_ET ); + cr0 = (cr0 & ~(mask & ~CR0_PG_MASK )); +wvmcs (vcpu,VMCS_GUEST_CR0 , cr0 |CR0_NE_MASK |CR0_ET_MASK );hv_vcpu_invalidate_tlb (vcpu);hv_vcpu_flush (vcpu);
@@ -164,11 +165,11 @@ static inline void macvm_set_cr0(hv_vcpuid_t vcpu, uint64_t cr0)static inline void macvm_set_cr4 (hv_vcpuid_t vcpu,uint64_t cr4) { -uint64_t guest_cr4 = cr4 |CR4_VMXE ; +uint64_t guest_cr4 = cr4 |CR4_VMXE_MASK ;wvmcs (vcpu,VMCS_GUEST_CR4 , guest_cr4);wvmcs (vcpu,VMCS_CR4_SHADOW , cr4); -wvmcs (vcpu,VMCS_CR4_MASK ,CR4_VMXE ); +wvmcs (vcpu,VMCS_CR4_MASK ,CR4_VMXE_MASK );hv_vcpu_invalidate_tlb (vcpu);hv_vcpu_flush (vcpu);
target/i386/hvf/x86.c +3 -3
@@ -119,7 +119,7 @@ bool x86_read_call_gate(struct CPUState *cpu, struct x86_call_gate *idt_desc,bool x86_is_protected (struct CPUState *cpu) {uint64_t cr0 =rvmcs (cpu->hvf->fd,VMCS_GUEST_CR0 ); -return cr0 &CR0_PE ; +return cr0 &CR0_PE_MASK ; }bool x86_is_real (struct CPUState *cpu)
@@ -150,13 +150,13 @@ bool x86_is_long64_mode(struct CPUState *cpu)bool x86_is_paging_mode (struct CPUState *cpu) {uint64_t cr0 =rvmcs (cpu->hvf->fd,VMCS_GUEST_CR0 ); -return cr0 &CR0_PG ; +return cr0 &CR0_PG_MASK ; }bool x86_is_pae_enabled (struct CPUState *cpu) {uint64_t cr4 =rvmcs (cpu->hvf->fd,VMCS_GUEST_CR4 ); -return cr4 &CR4_PAE ; +return cr4 &CR4_PAE_MASK ; }target_ulong linear_addr (struct CPUState *cpu,target_ulong addr,X86Seg seg)
target/i386/hvf/x86.h +0 -34
@@ -42,40 +42,6 @@ typedef struct x86_register { }; } __attribute__ ((__packed__)) x86_register; -typedef enum x86_reg_cr0 { -CR0_PE = (1L <<0 ), -CR0_MP = (1L <<1 ), -CR0_EM = (1L <<2 ), -CR0_TS = (1L <<3 ), -CR0_ET = (1L <<4 ), -CR0_NE = (1L <<5 ), -CR0_WP = (1L <<16 ), -CR0_AM = (1L <<18 ), -CR0_NW = (1L <<29 ), -CR0_CD = (1L <<30 ), -CR0_PG = (1L <<31 ), -}x86_reg_cr0 ; - -typedef enum x86_reg_cr4 { -CR4_VME = (1L <<0 ), -CR4_PVI = (1L <<1 ), -CR4_TSD = (1L <<2 ), -CR4_DE = (1L <<3 ), -CR4_PSE = (1L <<4 ), -CR4_PAE = (1L <<5 ), -CR4_MSE = (1L <<6 ), -CR4_PGE = (1L <<7 ), -CR4_PCE = (1L <<8 ), -CR4_OSFXSR = (1L <<9 ), -CR4_OSXMMEXCPT = (1L <<10 ), -CR4_VMXE = (1L <<13 ), -CR4_SMXE = (1L <<14 ), -CR4_FSGSBASE = (1L <<16 ), -CR4_PCIDE = (1L <<17 ), -CR4_OSXSAVE = (1L <<18 ), -CR4_SMEP = (1L <<20 ), -}x86_reg_cr4 ; -/* 16 bit Task State Segment */ typedef struct x86_tss_segment16 { uint16_t link;
target/i386/hvf/x86_mmu.c +1 -1
@@ -129,7 +129,7 @@ static bool test_pt_entry(struct CPUState *cpu, struct gpt_translation *pt,uint32_t cr0 =rvmcs (cpu->hvf->fd,VMCS_GUEST_CR0 );/* check protection */ -if (cr0 &CR0_WP ) { +if (cr0 &CR0_WP_MASK ) {if (pt->write_access && !pte_write_access (pte)) {return false; }
target/i386/hvf/x86_task.c +2 -1
@@ -174,7 +174,8 @@ void vmx_handle_task_switch(CPUState *cpu, x68_segment_selector tss_sel, int rea//ret = task_switch_16(cpu, tss_sel, old_tss_sel, old_tss_base, &next_tss_desc); VM_PANIC ("task_switch_16" ); -macvm_set_cr0 (cpu->hvf->fd,rvmcs (cpu->hvf->fd,VMCS_GUEST_CR0 ) |CR0_TS ); +macvm_set_cr0 (cpu->hvf->fd,rvmcs (cpu->hvf->fd,VMCS_GUEST_CR0 ) | +CR0_TS_MASK );x86_segment_descriptor_to_vmx (cpu, tss_sel, &next_tss_desc, &vmx_seg);vmx_write_segment_descriptor (cpu, &vmx_seg,R_TR );
Signature · 1 line
2.34.1
PATCH v7 4/22 hvf: Make hvf_get_segments() / hvf_put_segments() local
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Both hvf_get_segments/hvf_put_segments() functions are only used within x86hvf.c: do not declare them as public API.
--- target/i386/hvf/x86hvf.c | 4 ++-- target/i386/hvf/x86hvf.h | 2 -- 2 files changed, 2 insertions(+), 4 deletions(-)
target/i386/hvf/x86hvf.c +2 -2
@@ -83,7 +83,7 @@ void hvf_put_xsave(CPUState *cpu_state) } } -void hvf_put_segments (CPUState *cpu_state) +static void hvf_put_segments (CPUState *cpu_state) {CPUX86State *env = &X86_CPU (cpu_state)->env;struct vmx_segment seg;
@@ -166,7 +166,7 @@ void hvf_get_xsave(CPUState *cpu_state)x86_cpu_xrstor_all_areas (X86_CPU (cpu_state), xsave, xsave_len); } -void hvf_get_segments (CPUState *cpu_state) +static void hvf_get_segments (CPUState *cpu_state) {CPUX86State *env = &X86_CPU (cpu_state)->env;
target/i386/hvf/x86hvf.h +0 -2
@@ -26,11 +26,9 @@ void hvf_set_segment(struct CPUState *cpu, struct vmx_segment *vmx_seg, SegmentCache *qseg, bool is_tr);void hvf_get_segment (SegmentCache *qseg,struct vmx_segment *vmx_seg);void hvf_put_xsave (CPUState *cpu_state); -void hvf_put_segments (CPUState *cpu_state);void hvf_put_msrs (CPUState *cpu_state);void hvf_get_xsave (CPUState *cpu_state);void hvf_get_msrs (CPUState *cpu_state);void vmx_clear_int_window_exiting (CPUState *cpu); -void hvf_get_segments (CPUState *cpu_state);void vmx_update_tpr (CPUState *cpu);#endif
Signature · 1 line
2.34.1
PATCH v7 5/22 hvf: Remove deprecated hv_vcpu_flush() calls
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
When building on macOS 11 [*], we get:
In file included from ../target/i386/hvf/hvf.c:59:
../target/i386/hvf/vmx.h:174:5: error: 'hv_vcpu_flush' is deprecated: first deprecated in macOS 11.0 - This API has no effect and always returns HV_UNSUPPORTED [-Werror,-Wdeprecated-declarations]
hv_vcpu_flush(vcpu);
^
/Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/System/Library/Frameworks/Hypervisor.framework/Headers/hv.h:364:20: note: 'hv_vcpu_flush' has been explicitly marked deprecated here
extern hv_return_t hv_vcpu_flush(hv_vcpuid_t vcpu)
^Since this call "has no effect", simply remove it ¯\_(ツ)_/¯
Not very useful deprecation doc:
https://developer.apple.com/documentation/hypervisor/1441386-hv_vcpu_flush
[*] Also 10.15 (Catalina):
https://lore.kernel.org/qemu-devel/[email protected]/
--- target/i386/hvf/vmx.h | 2 -- target/i386/hvf/x86_task.c | 1 - target/i386/hvf/x86hvf.c | 2 -- 3 files changed, 5 deletions(-)
target/i386/hvf/vmx.h +0 -2
@@ -160,7 +160,6 @@ static inline void macvm_set_cr0(hv_vcpuid_t vcpu, uint64_t cr0)wvmcs (vcpu,VMCS_GUEST_CR0 , cr0 |CR0_NE_MASK |CR0_ET_MASK );hv_vcpu_invalidate_tlb (vcpu); -hv_vcpu_flush (vcpu); }static inline void macvm_set_cr4 (hv_vcpuid_t vcpu,uint64_t cr4)
@@ -172,7 +171,6 @@ static inline void macvm_set_cr4(hv_vcpuid_t vcpu, uint64_t cr4)wvmcs (vcpu,VMCS_CR4_MASK ,CR4_VMXE_MASK );hv_vcpu_invalidate_tlb (vcpu); -hv_vcpu_flush (vcpu); }static inline void macvm_set_rip (CPUState *cpu,uint64_t rip)
target/i386/hvf/x86_task.c +0 -1
@@ -182,5 +182,4 @@ void vmx_handle_task_switch(CPUState *cpu, x68_segment_selector tss_sel, int reastore_regs (cpu);hv_vcpu_invalidate_tlb (cpu->hvf->fd); -hv_vcpu_flush (cpu->hvf->fd); }
target/i386/hvf/x86hvf.c +0 -2
@@ -125,8 +125,6 @@ static void hvf_put_segments(CPUState *cpu_state)hvf_set_segment (cpu_state, &seg, &env->ldt, false);vmx_write_segment_descriptor (cpu_state, &seg,R_LDTR ); - -hv_vcpu_flush (cpu_state->hvf->fd); }void hvf_put_msrs (CPUState *cpu_state)
Signature · 1 line
2.34.1
PATCH v7 6/22 hvf: Fix OOB write in RDTSCP instruction decode
In reply to Philippe Mathieu-DaudéFrom: Cameron Esfahani <[email protected]>
A guest could craft a specific stream of instructions that will have QEMU write 0xF9 to inappropriate locations in memory. Add additional asserts to check for this. Generate a #UD if there are more than 14 prefix bytes.
Found by Julian Stecklina <[email protected]>
Message-Id: <[email protected]>
--- target/i386/hvf/x86_decode.c | 12 ++++++++++-- target/i386/hvf/x86hvf.c | 8 ++++++++ target/i386/hvf/x86hvf.h | 1 + 3 files changed, 19 insertions(+), 2 deletions(-)
target/i386/hvf/x86_decode.c +10 -2
@@ -24,8 +24,10 @@#include "vmx.h" #include "x86_mmu.h" #include "x86_descr.h" +#include "x86hvf.h" #define OPCODE_ESCAPE 0xf +#define X86_MAX_INSN_PREFIX_LENGTH (14)static void decode_invalid (CPUX86State *env,struct x86_decode *decode) {
@@ -541,7 +543,8 @@ static void decode_lidtgroup(CPUX86State *env, struct x86_decode *decode) }; decode->cmd = group[decode->modrm.reg];if (0xf9 == decode->modrm.modrm) { - decode->opcode[decode->len++] = decode->modrm.modrm; +VM_PANIC_ON (decode->opcode_len >=sizeof (decode->opcode)); + decode->opcode[decode->opcode_len++] = decode->modrm.modrm; decode->cmd =X86_DECODE_CMD_RDTSCP ; } }
@@ -1847,7 +1850,8 @@ void calc_modrm_operand(CPUX86State *env, struct x86_decode *decode,static void decode_prefix (CPUX86State *env,struct x86_decode *decode) { -while (1 ) { +/* At most X86_MAX_INSN_PREFIX_LENGTH prefix bytes. */ +for (int i =0 ; i <X86_MAX_INSN_PREFIX_LENGTH ; i++) {/* * REX prefix must come after legacy prefixes. * REX before legacy is ignored.
@@ -1892,6 +1896,8 @@ static void decode_prefix(CPUX86State *env, struct x86_decode *decode)return ; } } +/* Too many prefixes! Generate #UD. */ +hvf_inject_ud (env); }void set_addressing_size (CPUX86State *env,struct x86_decode *decode)
@@ -2090,11 +2096,13 @@ static void decode_opcodes(CPUX86State *env, struct x86_decode *decode)uint8_t opcode; opcode =decode_byte (env, decode); +VM_PANIC_ON (decode->opcode_len >=sizeof (decode->opcode)); decode->opcode[decode->opcode_len++] = opcode;if (opcode !=OPCODE_ESCAPE ) {decode_opcode_1 (env, decode, opcode); }else { opcode =decode_byte (env, decode); +VM_PANIC_ON (decode->opcode_len >=sizeof (decode->opcode)); decode->opcode[decode->opcode_len++] = opcode;decode_opcode_2 (env, decode, opcode); }
target/i386/hvf/x86hvf.c +8 -0
@@ -423,6 +423,14 @@ bool hvf_inject_interrupts(CPUState *cpu_state) & (CPU_INTERRUPT_INIT | CPU_INTERRUPT_TPR)); } +void hvf_inject_ud (CPUX86State *env) +{ + env->exception_nr =EXCP06_ILLOP ; + env->exception_injected =1 ; + env->has_error_code = false; + env->error_code =0 ; +} +int hvf_process_events (CPUState *cpu_state) {X86CPU *cpu =X86_CPU (cpu_state);
target/i386/hvf/x86hvf.h +1 -0
@@ -22,6 +22,7 @@int hvf_process_events (CPUState *);bool hvf_inject_interrupts (CPUState *); +void hvf_inject_ud (CPUX86State *);void hvf_set_segment (struct CPUState *cpu,struct vmx_segment *vmx_seg,SegmentCache *qseg,bool is_tr);void hvf_get_segment (SegmentCache *qseg,struct vmx_segment *vmx_seg);
Signature · 1 line
2.34.1
PATCH v7 18/22 ui/cocoa: add option to disable left-command forwarding to guest
In reply to Philippe Mathieu-DaudéFrom: Carwyn Ellis <[email protected]>
When switching between guest and host on a Mac using command-tab the command key is sent to the guest which can trigger functionality in the guest OS. Specifying left-command-key=off disables forwarding this key to the guest. Defaults to enabled.
Also updated the cocoa display documentation to reference the new left-command-key option along with the existing show-cursor option.
--- qapi/ui.json | 18 ++++++++++++++++++ qemu-options.hx | 12 ++++++++++++ ui/cocoa.m | 8 +++++++- 3 files changed, 37 insertions(+), 1 deletion(-)
qapi/ui.json +18 -0
@@ -1260,6 +1260,23 @@ { 'struct' : 'DisplayCurses', 'data' : { '*charset' : 'str' } } +## +# @DisplayCocoa: +# +# Cocoa display options. +# +# @left-command-key: Enable/disable forwarding of left command key to +# guest. Allows command-tab window switching on the +# host without sending this key to the guest when +# "off". Defaults to "on" +# +# Since: 7.0 +## +{ 'struct': 'DisplayCocoa', + 'data': { + '*left-command-key': 'bool' + } } + ## # @DisplayType: #
@@ -1338,6 +1355,7 @@ 'discriminator' : 'type', 'data' : { 'gtk': { 'type': 'DisplayGTK', 'if': 'CONFIG_GTK' }, + 'cocoa': { 'type': 'DisplayCocoa', 'if': 'CONFIG_COCOA' }, 'curses': { 'type': 'DisplayCurses', 'if': 'CONFIG_CURSES' }, 'egl-headless': { 'type': 'DisplayEGLHeadless', 'if': { 'all': ['CONFIG_OPENGL', 'CONFIG_GBM'] } },
qemu-options.hx +12 -0
@@ -1922,6 +1922,9 @@ DEF("display", HAS_ARG, QEMU_OPTION_display, #if defined(CONFIG_DBUS_DISPLAY) "-display dbus[,addr=<dbusaddr>]\n" " [,gl=on|core|es|off][,rendernode=<file>]\n" +#endif +#if defined(CONFIG_COCOA) + "-display cocoa[,show-cursor=on|off][,left-command-key=on|off]\n" #endif "-display none\n" " select display backend type\n"
@@ -2009,6 +2012,15 @@ SRST ``charset=CP850`` for IBM CP850 encoding. The default is ``CP437``. + ``cocoa`` + Display video output in a Cocoa window. Mac only. This interface + provides drop-down menus and other UI elements to configure and + control the VM during runtime. Valid parameters are: + + ``show-cursor=on|off`` : Force showing the mouse cursor + + ``left-command-key=on|off`` : Disable forwarding left command key to host + ``egl-headless[,rendernode=<file>]`` Offload all OpenGL operations to a local DRI device. For any graphical display, this display needs to be paired with either
ui/cocoa.m +7 -1
@@ -95,6 +95,7 @@ static DisplayChangeListener dcl = { }; static int last_buttons; static int cursor_hide = 1; +static int left_command_key_enabled = 1; static int gArgc; static char **gArgv;
@@ -853,7 +854,8 @@ QemuCocoaView *cocoaView; /* Don't pass command key changes to guest unless mouse is grabbed */ case kVK_Command: if (isMouseGrabbed && - !!(modifiers & NSEventModifierFlagCommand)) { + !!(modifiers & NSEventModifierFlagCommand) && + left_command_key_enabled) { [self toggleKey:Q_KEY_CODE_META_L]; } break;
@@ -2068,6 +2070,10 @@ static void cocoa_display_init(DisplayState *ds, DisplayOptions *opts) cursor_hide = 0; } + if (opts->u.cocoa.has_left_command_key && !opts->u.cocoa.left_command_key) { + left_command_key_enabled = 0; + } + // register vga output callbacks register_displaychangelistener(&dcl);
Signature · 1 line
2.34.1
PATCH v7 7/22 block/file-posix: Remove a deprecation warning on macOS 12
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
When building on macOS 12 we get:
block/file-posix.c:3335:18: warning: 'IOMasterPort' is deprecated: first deprecated in macOS 12.0 [-Wdeprecated-declarations]
kernResult = IOMasterPort( MACH_PORT_NULL, &masterPort );
^~~~~~~~~~~~
IOMainPortReplace by IOMainPort, redefining it to IOMasterPort if not available.
Reviewed by: Cameron Esfahani <[email protected]>
--- block/file-posix.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-)
block/file-posix.c +10 -4
@@ -3319,17 +3319,23 @@ BlockDriver bdrv_file = {#if defined(__APPLE__) && defined(__MACH__)static kern_return_t GetBSDPath (io_iterator_t mediaIterator,char *bsdPath,CFIndex maxPathSize,int flags); + +#if !defined(MAC_OS_VERSION_12_0 ) \ + || (MAC_OS_X_VERSION_MIN_REQUIRED <MAC_OS_VERSION_12_0 ) +#define IOMainPort IOMasterPort +#endif +static char *FindEjectableOpticalMedia (io_iterator_t *mediaIterator) {kern_return_t kernResult =KERN_FAILURE ; -mach_port_t masterPort; +mach_port_t mainPort;CFMutableDictionaryRef classesToMatch;const char *matching_array[] = {kIODVDMediaClass, kIOCDMediaClass};char *mediaType =NULL ; - kernResult =IOMasterPort (MACH_PORT_NULL , &masterPort ); + kernResult =IOMainPort (MACH_PORT_NULL , &mainPort);if (KERN_SUCCESS != kernResult ) { -printf ("IOMasterPort returned %d\n" , kernResult ); +printf ("IOMainPort returned %d\n" , kernResult); }int index;
@@ -3342,7 +3348,7 @@ static char *FindEjectableOpticalMedia(io_iterator_t *mediaIterator) }CFDictionarySetValue (classesToMatch,CFSTR (kIOMediaEjectableKey), kCFBooleanTrue); - kernResult =IOServiceGetMatchingServices (masterPort, classesToMatch, + kernResult =IOServiceGetMatchingServices (mainPort, classesToMatch, mediaIterator);if (kernResult !=KERN_SUCCESS ) { error_report("Note: IOServiceGetMatchingServices returned %d",
Signature · 1 line
2.34.1
PATCH v7 11/22 coreaudio: Always return 0 in handle_voice_change
In reply to Philippe Mathieu-DaudéFrom: Akihiko Odaki <[email protected]>
handle_voice_change() is a CoreAudio callback function as of CoreAudio type AudioObjectPropertyListenerProc, and for the latter MacOSX.sdk/System/ Library/Frameworks/CoreAudio.framework/Headers/AudioHardware.h says "The return value is currently unused and should always be 0.".
Message-Id: <[email protected]>
--- audio/coreaudio.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
audio/coreaudio.c +2 -4
@@ -545,7 +545,6 @@ static OSStatus handle_voice_change(const AudioObjectPropertyAddress *in_addresses, void *in_client_data) { - OSStatus status; coreaudioVoiceOut *core = in_client_data;qemu_mutex_lock_iothread ();
@@ -554,13 +553,12 @@ static OSStatus handle_voice_change(fini_out_device (core); } - status =init_out_device (core); -if (!status) { +if (!init_out_device (core)) {update_device_playback_state (core); }qemu_mutex_unlock_iothread (); -return status; +return 0 ; } static int coreaudio_init_out(HWVoiceOut *hw, struct audsettings *as,
Signature · 1 line
2.34.1
PATCH v7 8/22 audio/coreaudio: Remove a deprecation warning on macOS 12
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
When building on macOS 12 we get:
audio/coreaudio.c:50:5: error: 'kAudioObjectPropertyElementMaster' is deprecated: first deprecated in macOS 12.0 [-Werror,-Wdeprecated-declarations]
kAudioObjectPropertyElementMaster
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
kAudioObjectPropertyElementMain
/Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/System/Library/Frameworks/CoreAudio.framework/Headers/AudioHardwareBase.h:208:5: note: 'kAudioObjectPropertyElementMaster' has been explicitly marked deprecated here
kAudioObjectPropertyElementMaster API_DEPRECATED_WITH_REPLACEMENT("kAudioObjectPropertyElementMain", macos(10.0, 12.0), ios(2.0, 15.0), watchos(1.0, 8.0), tvos(9.0, 15.0)) = kAudioObjectPropertyElementMain
^Replace by kAudioObjectPropertyElementMain, redefining it to kAudioObjectPropertyElementMaster if not available.
--- audio/coreaudio.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-)
audio/coreaudio.c +11 -6
@@ -44,10 +44,15 @@ typedef struct coreaudioVoiceOut {bool enabled; } coreaudioVoiceOut; +#if !defined(MAC_OS_VERSION_12_0 ) \ + || (MAC_OS_X_VERSION_MIN_REQUIRED <MAC_OS_VERSION_12_0 ) +#define kAudioObjectPropertyElementMain kAudioObjectPropertyElementMaster +#endif +static const AudioObjectPropertyAddress voice_addr = { kAudioHardwarePropertyDefaultOutputDevice, kAudioObjectPropertyScopeGlobal, - kAudioObjectPropertyElementMaster + kAudioObjectPropertyElementMain };static OSStatus coreaudio_get_voice (AudioDeviceID *id)
@@ -69,7 +74,7 @@ static OSStatus coreaudio_get_framesizerange(AudioDeviceID id,AudioObjectPropertyAddress addr = { kAudioDevicePropertyBufferFrameSizeRange, kAudioDevicePropertyScopeOutput, - kAudioObjectPropertyElementMaster + kAudioObjectPropertyElementMain }; return AudioObjectGetPropertyData(id,
@@ -86,7 +91,7 @@ static OSStatus coreaudio_get_framesize(AudioDeviceID id, UInt32 *framesize)AudioObjectPropertyAddress addr = { kAudioDevicePropertyBufferFrameSize, kAudioDevicePropertyScopeOutput, - kAudioObjectPropertyElementMaster + kAudioObjectPropertyElementMain }; return AudioObjectGetPropertyData(id,
@@ -103,7 +108,7 @@ static OSStatus coreaudio_set_framesize(AudioDeviceID id, UInt32 *framesize)AudioObjectPropertyAddress addr = { kAudioDevicePropertyBufferFrameSize, kAudioDevicePropertyScopeOutput, - kAudioObjectPropertyElementMaster + kAudioObjectPropertyElementMain }; return AudioObjectSetPropertyData(id,
@@ -121,7 +126,7 @@ static OSStatus coreaudio_set_streamformat(AudioDeviceID id,AudioObjectPropertyAddress addr = { kAudioDevicePropertyStreamFormat, kAudioDevicePropertyScopeOutput, - kAudioObjectPropertyElementMaster + kAudioObjectPropertyElementMain }; return AudioObjectSetPropertyData(id,
@@ -138,7 +143,7 @@ static OSStatus coreaudio_get_isrunning(AudioDeviceID id, UInt32 *result)AudioObjectPropertyAddress addr = { kAudioDevicePropertyDeviceIsRunning, kAudioDevicePropertyScopeOutput, - kAudioObjectPropertyElementMaster + kAudioObjectPropertyElementMain }; return AudioObjectGetPropertyData(id,
Signature · 1 line
2.34.1
PATCH v7 19/22 ui/cocoa: release mouse when user switches away from QEMU window
In reply to Philippe Mathieu-DaudéFrom: Carwyn Ellis <[email protected]>
This resolves an issue where using command-tab to switch between QEMU and other windows on the host can leave the mouse pointer visible.
By releasing the mouse when the user switches away, the user must left click on the QEMU window when switching back in order to hide the pointer and return control to the guest.
This appraoch ensures that the calls to NSCursor hide and unhide are always balanced and thus work correctly when invoked.
--- ui/cocoa.m | 2 ++ 1 file changed, 2 insertions(+)
ui/cocoa.m +2 -0
@@ -1264,6 +1264,7 @@ QemuCocoaView *cocoaView; - (void) applicationWillResignActive: (NSNotification *)aNotification { COCOA_DEBUG("QemuCocoaAppController: applicationWillResignActive\n"); + [cocoaView ungrabMouse]; [cocoaView raiseAllKeys]; }
@@ -2066,6 +2067,7 @@ static void cocoa_display_init(DisplayState *ds, DisplayOptions *opts) [(QemuCocoaAppController *)[[NSApplication sharedApplication] delegate] toggleFullScreen: nil]; }); } + if (opts->has_show_cursor && opts->show_cursor) { cursor_hide = 0; }
Signature · 1 line
2.34.1
PATCH v7 9/22 audio/dbus: Fix building with modules on macOS
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
When configuring QEMU with --enable-modules we get on macOS:
--- stderr ---
Dependency ui-dbus cannot be satisfied
ui-dbus depends on pixman and opengl, so add these dependencies to audio-dbus.
--- audio/meson.build | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
audio/meson.build +1 -1
@@ -28,7 +28,7 @@ endforeach if dbus_display module_ss = ss.source_set() - module_ss.add(when: gio, if_true: files('dbusaudio.c')) + module_ss.add(when: [gio, pixman, opengl, 'CONFIG_GIO'], if_true: files('dbusaudio.c')) audio_modules += {'dbus': module_ss} endif
Signature · 1 line
2.34.1
PATCH v7 10/22 audio: Log context for audio bug
In reply to Philippe Mathieu-DaudéFrom: Akihiko Odaki <[email protected]>
Without this change audio_bug aborts when the bug condition is met, which discards following useful logs. Call abort after such logs.
Message-Id: <[email protected]>
--- audio/audio.c | 25 ++++++++++++------------- audio/audio_template.h | 27 ++++++++++++--------------- 2 files changed, 24 insertions(+), 28 deletions(-)
audio/audio.c +12 -13
@@ -117,7 +117,6 @@ int audio_bug (const char *funcname, int cond)AUD_log (NULL ,"I am sorry\n" ); }AUD_log (NULL ,"Context:\n" ); -abort (); }return cond;
@@ -138,7 +137,7 @@ static inline int audio_bits_to_index (int bits)default :audio_bug ("bits_to_index" ,1 );AUD_log (NULL ,"invalid bits %d\n" , bits); -return 0 ; +abort (); } }
@@ -156,7 +155,7 @@ void *audio_calloc (const char *funcname, int nmemb, size_t size)AUD_log (NULL ,"%s passed invalid arguments to audio_calloc\n" , funcname);AUD_log (NULL ,"nmemb=%d size=%zu (len=%zu)\n" , nmemb, size, len); -return NULL ; +abort (); }return g_malloc0 (len);
@@ -543,7 +542,7 @@ static size_t audio_pcm_hw_get_live_in(HWVoiceIn *hw)size_t live = hw->total_samples_captured -audio_pcm_hw_find_min_in (hw);if (audio_bug (__func__, live > hw->conv_buf->size)) {dolog ("live=%zu hw->conv_buf->size=%zu\n" , live, hw->conv_buf->size); -return 0 ; +abort (); }return live; }
@@ -581,7 +580,7 @@ static size_t audio_pcm_sw_read(SWVoiceIn *sw, void *buf, size_t size) }if (audio_bug (__func__, live > hw->conv_buf->size)) {dolog ("live_in=%zu hw->conv_buf->size=%zu\n" , live, hw->conv_buf->size); -return 0 ; +abort (); } rpos =audio_ring_posb (hw->conv_buf->pos, live, hw->conv_buf->size);
@@ -656,7 +655,7 @@ static size_t audio_pcm_hw_get_live_out (HWVoiceOut *hw, int *nb_live)if (audio_bug (__func__, live > hw->mix_buf->size)) {dolog ("live=%zu hw->mix_buf->size=%zu\n" , live, hw->mix_buf->size); -return 0 ; +abort (); }return live; }
@@ -706,7 +705,7 @@ static size_t audio_pcm_sw_write(SWVoiceOut *sw, void *buf, size_t size) live = sw->total_hw_samples_mixed;if (audio_bug (__func__, live > hwsamples)) {dolog ("live=%zu hw->mix_buf->size=%zu\n" , live, hwsamples); -return 0 ; +abort (); }if (live == hwsamples) {
@@ -998,7 +997,7 @@ static size_t audio_get_avail (SWVoiceIn *sw)if (audio_bug (__func__, live > sw->hw->conv_buf->size)) {dolog ("live=%zu sw->hw->conv_buf->size=%zu\n" , live, sw->hw->conv_buf->size); -return 0 ; +abort (); } ldebug (
@@ -1028,7 +1027,7 @@ static size_t audio_get_free(SWVoiceOut *sw)if (audio_bug (__func__, live > sw->hw->mix_buf->size)) {dolog ("live=%zu sw->hw->mix_buf->size=%zu\n" , live, sw->hw->mix_buf->size); -return 0 ; +abort (); } dead = sw->hw->mix_buf->size - live;
@@ -1170,7 +1169,7 @@ static void audio_run_out (AudioState *s)if (audio_bug (__func__, live > hw->mix_buf->size)) {dolog ("live=%zu hw->mix_buf->size=%zu\n" , live, hw->mix_buf->size); -continue ; +abort (); }if (hw->pending_disable && !nb_live) {
@@ -1203,7 +1202,7 @@ static void audio_run_out (AudioState *s)if (audio_bug (__func__, hw->mix_buf->pos >= hw->mix_buf->size)) {dolog ("hw->mix_buf->pos=%zu hw->mix_buf->size=%zu played=%zu\n" , hw->mix_buf->pos, hw->mix_buf->size, played); - hw->mix_buf->pos =0 ; +abort (); }#ifdef DEBUG_OUT
@@ -1223,7 +1222,7 @@ static void audio_run_out (AudioState *s)if (audio_bug (__func__, played > sw->total_hw_samples_mixed)) {dolog ("played=%zu sw->total_hw_samples_mixed=%zu\n" , played, sw->total_hw_samples_mixed); - played = sw->total_hw_samples_mixed; +abort (); } sw->total_hw_samples_mixed -= played;
@@ -1346,7 +1345,7 @@ static void audio_run_capture (AudioState *s)if (audio_bug (__func__, captured > sw->total_hw_samples_mixed)) {dolog ("captured=%zu sw->total_hw_samples_mixed=%zu\n" , captured, sw->total_hw_samples_mixed); - captured = sw->total_hw_samples_mixed; +abort (); } sw->total_hw_samples_mixed -= captured;
audio/audio_template.h +12 -15
@@ -59,12 +59,13 @@ static void glue(audio_init_nb_voices_, TYPE)(AudioState *s,if (audio_bug (__func__, !voice_size && max_voices)) {dolog ("drv=`%s' voice_size=0 max_voices=%d\n" , drv->name, max_voices); -glue (s->nb_hw_voices_,TYPE ) =0 ; +abort (); }if (audio_bug (__func__, voice_size && !max_voices)) {dolog ("drv=`%s' voice_size=%d max_voices=0\n" , drv->name, voice_size); +abort (); } }
@@ -81,6 +82,7 @@ static void glue(audio_pcm_hw_alloc_resources_, TYPE)(HW *hw)size_t samples = hw->samples;if (audio_bug (__func__, samples ==0 )) {dolog ("Attempted to allocate empty buffer\n" ); +abort (); }HWBUF =g_malloc0 (sizeof (STSampleBuffer) +sizeof (st_sample) * samples);
@@ -252,12 +254,12 @@ static HW *glue(audio_pcm_hw_add_new_, TYPE)(AudioState *s,if (audio_bug (__func__, !drv)) {dolog ("No host audio driver\n" ); -return NULL ; +abort (); }if (audio_bug (__func__, !drv->pcm_ops)) {dolog ("Host audio driver without pcm_ops\n" ); -return NULL ; +abort (); } hw =audio_calloc (__func__,1 ,glue (drv->voice_size_,TYPE ));
@@ -275,12 +277,13 @@ static HW *glue(audio_pcm_hw_add_new_, TYPE)(AudioState *s,QLIST_INIT (&hw->cap_head);#endif if (glue (hw->pcm_ops->init_,TYPE ) (hw, as, s->drv_opaque)) { - goto err0; +g_free (hw); +return NULL ; }if (audio_bug (__func__, hw->samples <=0 )) {dolog ("hw->samples=%zd\n" , hw->samples); - goto err1; +abort (); }if (hw->info.is_float) {
@@ -309,12 +312,6 @@ static HW *glue(audio_pcm_hw_add_new_, TYPE)(AudioState *s,audio_attach_capture (hw);#endif return hw; - - err1: -glue (hw->pcm_ops->fini_,TYPE ) (hw); - err0: -g_free (hw); -return NULL ; }AudiodevPerDirectionOptions *glue (audio_get_pdo_ ,TYPE )(Audiodev *dev)
@@ -435,7 +432,7 @@ void glue (AUD_close_, TYPE) (QEMUSoundCard *card, SW *sw)if (sw) {if (audio_bug (__func__, !card)) {dolog ("card=%p\n" , card); -return ; +abort (); }glue (audio_close_,TYPE ) (sw);
@@ -457,7 +454,7 @@ SW *glue (AUD_open_, TYPE) (if (audio_bug (__func__, !card || !name || !callback_fn || !as)) {dolog ("card=%p name=%p callback_fn=%p as=%p\n" , card, name, callback_fn, as); - goto fail; +abort (); } s = card->state;
@@ -468,12 +465,12 @@ SW *glue (AUD_open_, TYPE) (if (audio_bug (__func__,audio_validate_settings (as))) {audio_print_settings (as); - goto fail; +abort (); }if (audio_bug (__func__, !s->drv)) {dolog ("Can not open `%s' (no host audio driver)\n" , name); - goto fail; +abort (); }if (sw &&audio_pcm_info_eq (&sw->info, as)) {
Signature · 1 line
2.34.1
PATCH v7 12/22 audio: Rename coreaudio extension to use Objective-C compiler
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
The coreaudio library includes Objective-C declarations (using the caret '^' symbol to declare block references [*]). When building with a C compiler we get:
[175/839] Compiling C object libcommon.fa.p/audio_coreaudio.c.o
In file included from /Library/Developer/CommandLineTools/SDKs/MacOSX12.sdk/System/Library/Frameworks/CoreAudio.framework/Headers/CoreAudio.h:18,
from ../../audio/coreaudio.c:26:
/Library/Developer/CommandLineTools/SDKs/MacOSX12.sdk/System/Library/Frameworks/CoreAudio.framework/Headers/AudioHardware.h:162:2: error: expected identifier or '(' before '^' token
162 | (^AudioObjectPropertyListenerBlock)( UInt32 inNumberAddresses,
| ^
FAILED: libcommon.fa.p/audio_coreaudio.c.o
Rename the file to use the Objective-C default extension (.m) so meson calls the correct compiler.
---
audio/{coreaudio.c => coreaudio.m} | 0
audio/meson.build | 2 +-
2 files changed, 1 insertion(+), 1 deletion(-)
rename audio/{coreaudio.c => coreaudio.m} (100%)audio/coreaudio.m renamed from audio/coreaudio.c +0 -0
audio/meson.build +1 -1
@@ -7,7 +7,7 @@ softmmu_ss.add(files( 'wavcapture.c', )) -softmmu_ss.add(when: coreaudio, if_true: files('coreaudio.c')) +softmmu_ss.add(when: coreaudio, if_true: files('coreaudio.m')) softmmu_ss.add(when: dsound, if_true: files('dsoundaudio.c', 'audio_win_int.c')) audio_modules = {}
Signature · 1 line
2.34.1
PATCH v7 20/22 ui/cocoa: capture all keys and combos when mouse is grabbed
In reply to Philippe Mathieu-DaudéFrom: Gustavo Noronha Silva <[email protected]>
Applications such as Gnome may use Alt-Tab and Super-Tab for different purposes, some use Ctrl-arrows so we want to allow qemu to handle everything when it captures the mouse/keyboard.
However, Mac OS handles some combos like Command-Tab and Ctrl-arrows at an earlier part of the event handling chain, not letting qemu see it.
We add a global Event Tap that allows qemu to see all events when the mouse is grabbed. Note that this requires additional permissions.
See:
https://developer.apple.com/documentation/coregraphics/1454426-cgeventtapcreate?language=objc#discussion
https://support.apple.com/en-in/guide/mac-help/mh32356/mac
Message-Id: <[email protected]>
Message-Id: <[email protected]>
--- qapi/ui.json | 8 +++++- qemu-options.hx | 3 +++ ui/cocoa.m | 65 ++++++++++++++++++++++++++++++++++++++++++++++++- 3 files changed, 74 insertions(+), 2 deletions(-)
qapi/ui.json +7 -1
@@ -1270,11 +1270,17 @@ # host without sending this key to the guest when # "off". Defaults to "on" # +# @full-grab: Capture all key presses, including system combos. This +# requires accessibility permissions, since it performs +# a global grab on key events. (default: off) +# See https://support.apple.com/en-in/guide/mac-help/mh32356/mac +# # Since: 7.0 ## { 'struct': 'DisplayCocoa', 'data': { - '*left-command-key': 'bool' + '*left-command-key': 'bool', + '*full-grab': 'bool' } } ##
qemu-options.hx +3 -0
@@ -1916,6 +1916,9 @@ DEF("display", HAS_ARG, QEMU_OPTION_display, #if defined(CONFIG_CURSES) "-display curses[,charset=<encoding>]\n" #endif +#if defined(CONFIG_COCOA) + "-display cocoa[,full_grab=on|off]\n" +#endif #if defined(CONFIG_OPENGL) "-display egl-headless[,rendernode=<file>]\n" #endif
ui/cocoa.m +64 -1
110 lines not on this page · raw message
Signature · 1 line
2.34.1
PATCH v7 13/22 osdep: Avoid using Clang-specific __builtin_available()
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Remove the Clang specific __builtin_available() to allow building with GCC, otherwise we get:
include/qemu/osdep.h: In function 'qemu_thread_jit_write':
include/qemu/osdep.h:787:9: warning: implicit declaration of function '__builtin_available'; did you mean '__builtin_scalbl'? [-Wimplicit-function-declaration]
787 | if (__builtin_available(macOS 11.0, *)) {
| ^~~~~~~~~~~~~~~~~~~
| __builtin_scalbl
include/qemu/osdep.h:787:9: warning: nested extern declaration of '__builtin_available' [-Wnested-externs]
include/qemu/osdep.h:787:29: error: 'macOS' undeclared (first use in this function)
787 | if (__builtin_available(macOS 11.0, *)) {
| ^~~~~
include/qemu/osdep.h:787:29: note: each undeclared identifier is reported only once for each function it appears in
include/qemu/osdep.h:787:34: error: expected ')' before numeric constant
787 | if (__builtin_available(macOS 11.0, *)) {
| ~ ^~~~~
| )Beside, on macOS Catalina we get 2254 times:
include/qemu/osdep.h:780:5: warning: 'pthread_jit_write_protect_np' is only available on macOS 11.0 or newer [-Wunguarded-availability-new]
pthread_jit_write_protect_np(true);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~Fix by using a stricker toolchain version low range, replacing MAC_OS_X_VERSION_MAX_ALLOWED by MAC_OS_X_VERSION_MIN_REQUIRED.
--- include/qemu/osdep.h | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-)
include/qemu/osdep.h +3 -7
@@ -642,19 +642,15 @@ size_t qemu_get_host_physmem(void);* for the current thread. */ #if defined(MAC_OS_VERSION_11_0 ) && \ -MAC_OS_X_VERSION_MAX_ALLOWED >=MAC_OS_VERSION_11_0 +MAC_OS_X_VERSION_MIN_REQUIRED >=MAC_OS_VERSION_11_0 static inline void qemu_thread_jit_execute (void ) { - if (__builtin_available(macOS 11.0, *)) { -pthread_jit_write_protect_np (true); - } +pthread_jit_write_protect_np (true); }static inline void qemu_thread_jit_write (void ) { - if (__builtin_available(macOS 11.0, *)) { -pthread_jit_write_protect_np (false); - } +pthread_jit_write_protect_np (false); }#else static inline void qemu_thread_jit_write (void ) {}
Signature · 1 line
2.34.1
PATCH v7 15/22 meson: Log QEMU_CXXFLAGS content in summary
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
--- meson.build | 1 + 1 file changed, 1 insertion(+)
meson.build +1 -0
@@ -3470,6 +3470,7 @@ if link_args.length() > 0 summary_info += {'LDFLAGS': ' '.join(link_args)} endif summary_info += {'QEMU_CFLAGS': config_host['QEMU_CFLAGS']} +summary_info += {'QEMU_CXXFLAGS': config_host['QEMU_CXXFLAGS']} summary_info += {'QEMU_LDFLAGS': config_host['QEMU_LDFLAGS']} summary_info += {'profiler': get_option('profiler')} summary_info += {'link-time optimization (LTO)': get_option('b_lto')}
Signature · 1 line
2.34.1
PATCH v7 22/22 gitlab-ci: Support macOS 12 via cirrus-run
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Add support for macOS 12 build on Cirrus-CI, similarly to commit 0e103a65ba1 ("gitlab: support for ... macOS 11 via cirrus-run"), but with the following differences:
- Enable modules (configure --enable-modules)
- Do not run softfloat3 tests (make check-softfloat)
- Run Aarch64 qtests instead of x86_64 ones
Generate the vars file by calling 'make lcitool-refresh'.
--- .gitlab-ci.d/cirrus.yml | 16 ++++++++++++++++ .gitlab-ci.d/cirrus/macos-12.vars | 16 ++++++++++++++++ tests/lcitool/refresh | 1 + 3 files changed, 33 insertions(+) create mode 100644 .gitlab-ci.d/cirrus/macos-12.vars
.gitlab-ci.d/cirrus.yml +16 -0
@@ -87,6 +87,22 @@ x64-macos-11-base-build: PKG_CONFIG_PATH: /usr/local/opt/curl/lib/pkgconfig:/usr/local/opt/ncurses/lib/pkgconfig:/usr/local/opt/readline/lib/pkgconfig TEST_TARGETS: check-unit check-block check-qapi-schema check-softfloat check-qtest-x86_64 +x64-macos-12-base-build: + extends: .cirrus_build_job + variables: + NAME: macos-12 + CIRRUS_VM_INSTANCE_TYPE: osx_instance + CIRRUS_VM_IMAGE_SELECTOR: image + CIRRUS_VM_IMAGE_NAME: monterey-base + CIRRUS_VM_CPUS: 12 + CIRRUS_VM_RAM: 24G + UPDATE_COMMAND: brew update + INSTALL_COMMAND: brew install + PATH_EXTRA: /usr/local/opt/ccache/libexec:/usr/local/opt/gettext/bin + PKG_CONFIG_PATH: /usr/local/opt/curl/lib/pkgconfig:/usr/local/opt/ncurses/lib/pkgconfig:/usr/local/opt/readline/lib/pkgconfig + CONFIGURE_ARGS: --enable-modules + TEST_TARGETS: check-unit check-block check-qapi-schema check-qtest-aarch64 + # The following jobs run VM-based tests via KVM on a Linux-based Cirrus-CI job .cirrus_kvm_job:
.gitlab-ci.d/cirrus/macos-12.vars new file +16 -0
@@ -0,0 +1,16 @@ +# THIS FILE WAS AUTO-GENERATED +# +# $ lcitool variables macos-12 qemu +# +# https://gitlab.com/libvirt/libvirt-ci + +CCACHE='/usr/local/bin/ccache' +CPAN_PKGS='Test::Harness' +CROSS_PKGS='' +MAKE='/usr/local/bin/gmake' +NINJA='/usr/local/bin/ninja' +PACKAGING_COMMAND='brew' +PIP3='/usr/local/bin/pip3' +PKGS='bash bc bzip2 capstone ccache cpanminus ctags curl dbus diffutils dtc gcovr gettext git glib gnu-sed gnutls gtk+3 jemalloc jpeg-turbo libepoxy libffi libgcrypt libiscsi libnfs libpng libslirp libssh libtasn1 libusb llvm lzo make meson ncurses nettle ninja perl pixman pkg-config python3 rpm2cpio sdl2 sdl2_image snappy sparse spice-protocol tesseract texinfo usbredir vde vte3 zlib zstd' +PYPI_PKGS='PyYAML numpy pillow sphinx sphinx-rtd-theme virtualenv' +PYTHON='/usr/local/bin/python3'
tests/lcitool/refresh +1 -0
@@ -105,6 +105,7 @@ try: generate_cirrus("freebsd-12") generate_cirrus("freebsd-13") generate_cirrus("macos-11") + generate_cirrus("macos-12") sys.exit(0) except Exception as ex:
Signature · 1 line
2.34.1
PATCH v7 14/22 meson: Resolve the entitlement.sh script once for good
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Commit 235b523dba ("meson: Use find_program() to resolve the entitlement.sh script") didn't correctly fixed the issue, as the script is still resolved for each target. Move the check earlier, before processing each target.
--- meson.build | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
meson.build +4 -1
@@ -3063,6 +3063,10 @@ common_all = static_library('common', feature_to_c = find_program('scripts/feature_to_c.sh') +if targetos == 'darwin' + entitlement = find_program('scripts/entitlement.sh') +endif + emulators = {} foreach target : target_dirs config_target = config_target_mak[target]
@@ -3220,7 +3224,6 @@ foreach target : target_dirs install_input += meson.current_source_dir() / entitlements endif - entitlement = find_program('scripts/entitlement.sh') emulators += {exe['name'] : custom_target(exe['name'], input: build_input, output: exe['name'],
Signature · 1 line
2.34.1
PATCH v7 21/22 ui/cocoa: add option to swap Option and Command
In reply to Philippe Mathieu-DaudéFrom: Gustavo Noronha Silva <[email protected]>
On Mac OS X the Option key maps to Alt and Command to Super/Meta. This change swaps them around so that Alt is the key closer to the space bar and Meta/Super is between Control and Alt, like on non-Mac keyboards.
It is a cocoa display option, disabled by default.
Message-Id: <[email protected]>
Message-Id: <[email protected]>
--- qapi/ui.json | 7 ++++++- qemu-options.hx | 2 +- ui/cocoa.m | 49 +++++++++++++++++++++++++++++++++++++++---------- 3 files changed, 46 insertions(+), 12 deletions(-)
qapi/ui.json +6 -1
@@ -1275,12 +1275,17 @@ # a global grab on key events. (default: off) # See https://support.apple.com/en-in/guide/mac-help/mh32356/mac # +# @swap-opt-cmd: Swap the Option and Command keys so that their key codes match +# their position on non-Mac keyboards and you can use Meta/Super +# and Alt where you expect them. (default: off) +# # Since: 7.0 ## { 'struct': 'DisplayCocoa', 'data': { '*left-command-key': 'bool', - '*full-grab': 'bool' + '*full-grab': 'bool', + '*swap-opt-cmd': 'bool' } } ##
qemu-options.hx +1 -1
@@ -1917,7 +1917,7 @@ DEF("display", HAS_ARG, QEMU_OPTION_display, "-display curses[,charset=<encoding>]\n" #endif #if defined(CONFIG_COCOA) - "-display cocoa[,full_grab=on|off]\n" + "-display cocoa[,full-grab=on|off][,swap-opt-cmd=on|off]\n" #endif #if defined(CONFIG_OPENGL) "-display egl-headless[,rendernode=<file>]\n"
ui/cocoa.m +39 -10
92 lines not on this page · raw message
Signature · 1 line
2.34.1
PATCH v7 17/22 ui/cocoa: Constify qkeycode translation arrays
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
--- ui/cocoa.m | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
ui/cocoa.m +2 -2
@@ -671,7 +671,7 @@ QemuCocoaView *cocoaView; /* translates Macintosh keycodes to QEMU's keysym */ - int without_control_translation[] = { + static const int without_control_translation[] = { [0 ... 0xff] = 0, // invalid key [kVK_UpArrow] = QEMU_KEY_UP,
@@ -686,7 +686,7 @@ QemuCocoaView *cocoaView; [kVK_Delete] = QEMU_KEY_BACKSPACE, }; - int with_control_translation[] = { + static const int with_control_translation[] = { [0 ... 0xff] = 0, // invalid key [kVK_UpArrow] = QEMU_KEY_CTRL_UP,
Signature · 1 line
2.34.1
PATCH v7 16/22 configure: Pass filtered QEMU_OBJCFLAGS to meson
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Filter unsupported Objective-C options, to avoid 'unknown-warning-option' warnings when using Clang:
[34/373] Compiling Objective-C object libcommon.fa.p/audio_coreaudio.m.o
warning: unknown warning option '-Wold-style-declaration'; did you mean '-Wout-of-line-declaration'? [-Wunknown-warning-option]
warning: unknown warning option '-Wimplicit-fallthrough=2'; did you mean '-Wimplicit-fallthrough'? [-Wunknown-warning-option] 2 warnings generated.
--- configure | 23 +++++++++++++++++++++++ meson.build | 6 +++++- 2 files changed, 28 insertions(+), 1 deletion(-)
configure +23 -0
@@ -77,6 +77,7 @@ TMPB="qemu-conf" TMPC="${TMPDIR1}/${TMPB}.c" TMPO="${TMPDIR1}/${TMPB}.o" TMPCXX="${TMPDIR1}/${TMPB}.cxx" +TMPM="${TMPDIR1}/${TMPB}.m" TMPE="${TMPDIR1}/${TMPB}.exe" rm -f config.log
@@ -148,6 +149,10 @@ do_cxx() { do_compiler "$cxx" $CPU_CFLAGS "$@" } +do_objc() { + do_compiler "$objcc" $CPU_CFLAGS "$@" +} + # Append $2 to the variable named $1, with space separation add_to() { eval $1=\${$1:+\"\$$1 \"}\$2
@@ -1445,10 +1450,27 @@ cc_has_warning_flag() { compile_prog "-Werror $optflag" "" } +objcc_has_warning_flag() { + cat > $TMPM <<EOF +int main(void) { return 0; } +EOF + + # Use the positive sense of the flag when testing for -Wno-wombat + # support (gcc will happily accept the -Wno- form of unknown + # warning options). + optflag="$(echo $1 | sed -e 's/^-Wno-/-W/')" + do_objc -Werror $optflag \ + $OBJCFLAGS $EXTRA_OBJCFLAGS $CONFIGURE_OBJCFLAGS $QEMU_OBJCFLAGS \ + -o $TMPE $TMPM $QEMU_LDFLAGS +} + for flag in $gcc_flags; do if cc_has_warning_flag $flag ; then QEMU_CFLAGS="$QEMU_CFLAGS $flag" fi + if objcc_has_warning_flag $flag ; then + QEMU_OBJCFLAGS="$QEMU_OBJCFLAGS $flag" + fi done if test "$stack_protector" != "no"; then
@@ -2990,6 +3012,7 @@ echo "LD=$ld" >> $config_host_mak echo "CFLAGS_NOPIE=$CFLAGS_NOPIE" >> $config_host_mak echo "QEMU_CFLAGS=$QEMU_CFLAGS" >> $config_host_mak echo "QEMU_CXXFLAGS=$QEMU_CXXFLAGS" >> $config_host_mak +echo "QEMU_OBJCFLAGS=$QEMU_OBJCFLAGS" >> $config_host_mak echo "GLIB_CFLAGS=$glib_cflags" >> $config_host_mak echo "GLIB_LIBS=$glib_libs" >> $config_host_mak echo "GLIB_VERSION=$(pkg-config --modversion glib-2.0)" >> $config_host_mak
meson.build +5 -1
@@ -173,11 +173,13 @@ endif qemu_cflags = config_host['QEMU_CFLAGS'].split() qemu_cxxflags = config_host['QEMU_CXXFLAGS'].split() +qemu_objcflags = config_host['QEMU_OBJCFLAGS'].split() qemu_ldflags = config_host['QEMU_LDFLAGS'].split() if get_option('gprof') qemu_cflags += ['-p'] qemu_cxxflags += ['-p'] + qemu_objcflags += ['-p'] qemu_ldflags += ['-p'] endif
@@ -216,8 +218,9 @@ if get_option('fuzzing') endif endif -add_global_arguments(qemu_cflags, native: false, language: ['c', 'objc']) +add_global_arguments(qemu_cflags, native: false, language: ['c']) add_global_arguments(qemu_cxxflags, native: false, language: ['cpp']) +add_global_arguments(qemu_objcflags, native: false, language: ['objc']) add_global_link_arguments(qemu_ldflags, native: false, language: ['c', 'cpp', 'objc']) if targetos == 'linux'
@@ -3471,6 +3474,7 @@ if link_args.length() > 0 endif summary_info += {'QEMU_CFLAGS': config_host['QEMU_CFLAGS']} summary_info += {'QEMU_CXXFLAGS': config_host['QEMU_CXXFLAGS']} +summary_info += {'QEMU_OBJCFLAGS': config_host['QEMU_OBJCFLAGS']} summary_info += {'QEMU_LDFLAGS': config_host['QEMU_LDFLAGS']} summary_info += {'profiler': get_option('profiler')} summary_info += {'link-time optimization (LTO)': get_option('b_lto')}
Signature · 1 line
2.34.1
Re: [PATCH v7 13/22] osdep: Avoid using Clang-specific __builtin_available()
In reply to Philippe Mathieu-DaudéQuoted 65 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Remove the Clang specific
__builtin_available()to allow building with GCC, otherwise we get:
include/qemu/osdep.h: In function 'qemu_thread_jit_write':
include/qemu/osdep.h:787:9: warning: implicit declaration of function '__builtin_available'; did you mean '__builtin_scalbl'? [-Wimplicit-function-declaration]787 | if (__builtin_available(macOS 11.0, *)) { | ^~~~~~~~~~~~~~~~~~~ | __builtin_scalbl include/qemu/osdep.h:787:9: warning: nested extern declaration of '__builtin_available' [-Wnested-externs] include/qemu/osdep.h:787:29: error: 'macOS' undeclared (first use in this function) 787 | if (__builtin_available(macOS 11.0, *)) { | ^~~~~ include/qemu/osdep.h:787:29: note: each undeclared identifier is reported only once for each function it appears in include/qemu/osdep.h:787:34: error: expected ')' before numeric constant 787 | if (__builtin_available(macOS 11.0, *)) { | ~ ^~~~~ | )Beside, on macOS Catalina we get 2254 times:
include/qemu/osdep.h:780:5: warning: 'pthread_jit_write_protect_np' is only available on macOS 11.0 or newer [-Wunguarded-availability-new]pthread_jit_write_protect_np(true); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~Fix by using a stricker toolchain version low range, replacing
MAC_OS_X_VERSION_MAX_ALLOWEDbyMAC_OS_X_VERSION_MIN_REQUIRED.Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- include/qemu/osdep.h | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-)include/qemu/osdep.h +3
-8@@ -642,19 +642,15 @@ size_t qemu_get_host_physmem(void);* for the current thread. */ #if defined(MAC_OS_VERSION_11_0 ) && \ -MAC_OS_X_VERSION_MAX_ALLOWED >=MAC_OS_VERSION_11_0 +MAC_OS_X_VERSION_MIN_REQUIRED >=MAC_OS_VERSION_11_0 static inline void qemu_thread_jit_execute (void ) { - if (__builtin_available(macOS 11.0, *)) { -pthread_jit_write_protect_np (true); - } +pthread_jit_write_protect_np (true); }static inline void qemu_thread_jit_write (void ) { - if (__builtin_available(macOS 11.0, *)) { -pthread_jit_write_protect_np (false); - } +pthread_jit_write_protect_np (false); }#else static inline void qemu_thread_jit_write (void ) {} --2.34.1
Re: [PATCH v7 14/22] meson: Resolve the entitlement.sh script once for good
In reply to Philippe Mathieu-DaudéQuoted 39 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Commit
235b523dba("meson: Usefind_program()to resolve theentitlement.shscript") didn't correctly fixed the issue, as the script is still resolved for each target. Move the check earlier, before processing each target.Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- meson.build | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)meson.build +4
-2@@ -3063,6 +3063,10 @@ common_all = static_library('common', feature_to_c = find_program('scripts/feature_to_c.sh') +if targetos == 'darwin' + entitlement = find_program('scripts/entitlement.sh') +endif + emulators = {} foreach target : target_dirs config_target = config_target_mak[target]@@ -3220,7 +3224,6 @@ foreach target : target_dirs install_input += meson.current_source_dir() / entitlements endif - entitlement = find_program('scripts/entitlement.sh') emulators += {exe['name'] : custom_target(exe['name'], input: build_input, output: exe['name'], --2.34.1
Re: [PATCH v7 15/22] meson: Log QEMU_CXXFLAGS content in summary
In reply to Philippe Mathieu-DaudéQuoted 23 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- meson.build | 1 + 1 file changed, 1 insertion(+)meson.build +1
-1@@ -3470,6 +3470,7 @@ if link_args.length() > 0 summary_info += {'LDFLAGS': ' '.join(link_args)} endif summary_info += {'QEMU_CFLAGS': config_host['QEMU_CFLAGS']} +summary_info += {'QEMU_CXXFLAGS': config_host['QEMU_CXXFLAGS']} summary_info += {'QEMU_LDFLAGS': config_host['QEMU_LDFLAGS']} summary_info += {'profiler': get_option('profiler')} summary_info += {'link-time optimization (LTO)': get_option('b_lto')} --2.34.1
Re: [PATCH v7 16/22] configure: Pass filtered QEMU_OBJCFLAGS to meson
In reply to Philippe Mathieu-DaudéQuoted 117 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Filter unsupported Objective-C options, to avoid 'unknown-warning-option' warnings when using Clang:
[34/373] Compiling Objective-C object
libcommon.fa.p/audio_coreaudio.m.o
warning: unknown warning option '-Wold-style-declaration'; did you mean '-Wout-of-line-declaration'? [-Wunknown-warning-option]
warning: unknown warning option '-Wimplicit-fallthrough=2'; did you mean '-Wimplicit-fallthrough'? [-Wunknown-warning-option] 2 warnings generated.Acked-by Paolo Bonzini <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- configure | 23 +++++++++++++++++++++++ meson.build | 6 +++++- 2 files changed, 28 insertions(+), 1 deletion(-)configure +23
-0@@ -77,6 +77,7 @@ TMPB="qemu-conf" TMPC="${TMPDIR1}/${TMPB}.c" TMPO="${TMPDIR1}/${TMPB}.o" TMPCXX="${TMPDIR1}/${TMPB}.cxx" +TMPM="${TMPDIR1}/${TMPB}.m" TMPE="${TMPDIR1}/${TMPB}.exe" rm -f config.log@@ -148,6 +149,10 @@ do_cxx() { do_compiler "$cxx" $CPU_CFLAGS "$@" } +do_objc() { + do_compiler "$objcc" $CPU_CFLAGS "$@" +} + # Append $2 to the variable named $1, with space separation add_to() { eval $1=\${$1:+\"\$$1 \"}\$2@@ -1445,10 +1450,27 @@ cc_has_warning_flag() { compile_prog "-Werror $optflag" "" } +objcc_has_warning_flag() { + cat > $TMPM <<EOF +int main(void) { return 0; } +EOF + + # Use the positive sense of the flag when testing for -Wno-wombat + # support (gcc will happily accept the -Wno- form of unknown + # warning options). + optflag="$(echo $1 | sed -e 's/^-Wno-/-W/')" + do_objc -Werror $optflag \ + $OBJCFLAGS $EXTRA_OBJCFLAGS $CONFIGURE_OBJCFLAGS $QEMU_OBJCFLAGS \ + -o $TMPE $TMPM $QEMU_LDFLAGS +} + for flag in $gcc_flags; do if cc_has_warning_flag $flag ; then QEMU_CFLAGS="$QEMU_CFLAGS $flag" fi + if objcc_has_warning_flag $flag ; then + QEMU_OBJCFLAGS="$QEMU_OBJCFLAGS $flag" + fi done if test "$stack_protector" != "no"; then@@ -2990,6 +3012,7 @@ echo "LD=$ld" >> $config_host_mak echo "CFLAGS_NOPIE=$CFLAGS_NOPIE" >> $config_host_mak echo "QEMU_CFLAGS=$QEMU_CFLAGS" >> $config_host_mak echo "QEMU_CXXFLAGS=$QEMU_CXXFLAGS" >> $config_host_mak +echo "QEMU_OBJCFLAGS=$QEMU_OBJCFLAGS" >> $config_host_mak echo "GLIB_CFLAGS=$glib_cflags" >> $config_host_mak echo "GLIB_LIBS=$glib_libs" >> $config_host_mak echo "GLIB_VERSION=$(pkg-config --modversion glib-2.0)" >> $config_host_makmeson.build +5
-2@@ -173,11 +173,13 @@ endif qemu_cflags = config_host['QEMU_CFLAGS'].split() qemu_cxxflags = config_host['QEMU_CXXFLAGS'].split() +qemu_objcflags = config_host['QEMU_OBJCFLAGS'].split() qemu_ldflags = config_host['QEMU_LDFLAGS'].split() if get_option('gprof') qemu_cflags += ['-p'] qemu_cxxflags += ['-p'] + qemu_objcflags += ['-p'] qemu_ldflags += ['-p'] endif@@ -216,8 +218,9 @@ if get_option('fuzzing') endif endif -add_global_arguments(qemu_cflags, native: false, language: ['c', 'objc']) +add_global_arguments(qemu_cflags, native: false, language: ['c']) add_global_arguments(qemu_cxxflags, native: false, language: ['cpp']) +add_global_arguments(qemu_objcflags, native: false, language: ['objc']) add_global_link_arguments(qemu_ldflags, native: false, language: ['c', 'cpp', 'objc']) if targetos == 'linux'@@ -3471,6 +3474,7 @@ if link_args.length() > 0 endif summary_info += {'QEMU_CFLAGS': config_host['QEMU_CFLAGS']} summary_info += {'QEMU_CXXFLAGS': config_host['QEMU_CXXFLAGS']} +summary_info += {'QEMU_OBJCFLAGS': config_host['QEMU_OBJCFLAGS']} summary_info += {'QEMU_LDFLAGS': config_host['QEMU_LDFLAGS']} summary_info += {'profiler': get_option('profiler')} summary_info += {'link-time optimization (LTO)': get_option('b_lto')} --2.34.1
Re: [PATCH v7 17/22] ui/cocoa: Constify qkeycode translation arrays
In reply to Philippe Mathieu-DaudéQuoted 34 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Reported-by Peter Maydell <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- ui/cocoa.m | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)ui/cocoa.m +2
-3@@ -671,7 +671,7 @@ QemuCocoaView *cocoaView; /* translates Macintosh keycodes to QEMU's keysym */ - int without_control_translation[] = { + static const int without_control_translation[] = { [0 ... 0xff] = 0, // invalid key [kVK_UpArrow] = QEMU_KEY_UP,@@ -686,7 +686,7 @@ QemuCocoaView *cocoaView; [kVK_Delete] = QEMU_KEY_BACKSPACE, }; - int with_control_translation[] = { + static const int with_control_translation[] = { [0 ... 0xff] = 0, // invalid key [kVK_UpArrow] = QEMU_KEY_CTRL_UP, --2.34.1
Re: [PATCH v7 21/22] ui/cocoa: add option to swap Option and Command
In reply to Philippe Mathieu-DaudéQuoted 170 lines · Philippe Mathieu-Daudé
From: Gustavo Noronha Silva <[email protected]>
On Mac OS X the Option key maps to Alt and Command to Super/Meta. This change
swaps them around so that Alt is the key closer to the space bar and Meta/Super
is between Control and Alt, like on non-Mac keyboards.It is a cocoa display option, disabled by default.
Signed-off-by Gustavo Noronha Silva <[email protected]>Message-Id: <[email protected]>
Signed-off-by Akihiko Odaki <[email protected]>Message-Id: <[email protected]>
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- qapi/ui.json | 7 ++++++- qemu-options.hx | 2 +- ui/cocoa.m | 49 +++++++++++++++++++++++++++++++++++++++---------- 3 files changed, 46 insertions(+), 12 deletions(-)qapi/ui.json +0
-0@@ -1275,12 +1275,17 @@ # a global grab on key events. (default: off) # Seehttps://support.apple.com/en-in/guide/mac-help/mh32356/mac
#
+# @swap-opt-cmd: Swap the Option and Command keys so that their key codes match
+# their position on non-Mac keyboards and you can use
Meta/Superqapi/ui.json +4
-1+# and Alt where you expect them. (default: off) +# # Since: 7.0 ## { 'struct': 'DisplayCocoa', 'data': { '*left-command-key': 'bool', - '*full-grab': 'bool' + '*full-grab': 'bool', + '*swap-opt-cmd': 'bool' } } ##qemu-options.hx +1
-1@@ -1917,7 +1917,7 @@ DEF("display", HAS_ARG, QEMU_OPTION_display, "-display curses[,charset=<encoding>]\n" #endif #if defined(CONFIG_COCOA) - "-display cocoa[,full_grab=on|off]\n" + "-display cocoa[,full-grab=on|off][,swap-opt-cmd=on|off]\n" #endif #if defined(CONFIG_OPENGL) "-display egl-headless[,rendernode=<file>]\n"ui/cocoa.m +1
-0@@ -96,6 +96,7 @@ static DisplayChangeListener dcl = { static int last_buttons; static int cursor_hide = 1; static int left_command_key_enabled = 1; +static bool swap_opt_cmd; static int gArgc; static char **gArgv;@@ -854,12 +855,22 @@ static CGEventRef handleTapEvent(CGEventTapProxyproxy, CGEventType type, CGEven
qkbd_state_key_event(kbd, Q_KEY_CODE_CTRL_R, false); } if (!(modifiers & NSEventModifierFlagOption)) {ui/cocoa.m +14
-4- qkbd_state_key_event(kbd, Q_KEY_CODE_ALT, false); - qkbd_state_key_event(kbd, Q_KEY_CODE_ALT_R, false); + if (swap_opt_cmd) { + qkbd_state_key_event(kbd, Q_KEY_CODE_META_L, false); + qkbd_state_key_event(kbd, Q_KEY_CODE_META_R, false); + } else { + qkbd_state_key_event(kbd, Q_KEY_CODE_ALT, false); + qkbd_state_key_event(kbd, Q_KEY_CODE_ALT_R, false); + } } if (!(modifiers & NSEventModifierFlagCommand)) { - qkbd_state_key_event(kbd, Q_KEY_CODE_META_L, false); - qkbd_state_key_event(kbd, Q_KEY_CODE_META_R, false); + if (swap_opt_cmd) { + qkbd_state_key_event(kbd, Q_KEY_CODE_ALT, false); + qkbd_state_key_event(kbd, Q_KEY_CODE_ALT_R, false); + } else { + qkbd_state_key_event(kbd, Q_KEY_CODE_META_L, false); + qkbd_state_key_event(kbd, Q_KEY_CODE_META_R, false); + } } switch ([event type]) {@@ -891,29 +902,44 @@ static CGEventRef handleTapEvent(CGEventTapProxyproxy, CGEventType type, CGEven
case kVK_Option: if (!!(modifiers & NSEventModifierFlagOption)) {ui/cocoa.m +10
-2- [self toggleKey:Q_KEY_CODE_ALT]; + if (swap_opt_cmd) { + [self toggleKey:Q_KEY_CODE_META_L]; + } else { + [self toggleKey:Q_KEY_CODE_ALT]; + } } break; case kVK_RightOption: if (!!(modifiers & NSEventModifierFlagOption)) { - [self toggleKey:Q_KEY_CODE_ALT_R]; + if (swap_opt_cmd) { + [self toggleKey:Q_KEY_CODE_META_R]; + } else { + [self toggleKey:Q_KEY_CODE_ALT_R]; + } } break; /* Don't pass command key changes to guest unless mouseis grabbed */
case kVK_Command: if (isMouseGrabbed &&ui/cocoa.m +14
-4- !!(modifiers & NSEventModifierFlagCommand) && - left_command_key_enabled) { - [self toggleKey:Q_KEY_CODE_META_L]; + !!(modifiers & NSEventModifierFlagCommand)) { + if (swap_opt_cmd) { + [self toggleKey:Q_KEY_CODE_ALT]; + } else { + [self toggleKey:Q_KEY_CODE_META_L]; + } } break; case kVK_RightCommand: if (isMouseGrabbed && !!(modifiers & NSEventModifierFlagCommand)) { - [self toggleKey:Q_KEY_CODE_META_R]; + if (swap_opt_cmd) { + [self toggleKey:Q_KEY_CODE_ALT_R]; + } else { + [self toggleKey:Q_KEY_CODE_META_R]; + } } break; }@@ -2134,6 +2160,9 @@ static void cocoa_display_init(DisplayState *ds, DisplayOptions *opts) if (opts->has_show_cursor && opts->show_cursor) { cursor_hide = 0; } + if (opts->u.cocoa.has_swap_opt_cmd) { + swap_opt_cmd = opts->u.cocoa.swap_opt_cmd; + } if (opts->u.cocoa.has_left_command_key &&!opts->u.cocoa.left_command_key) { left_command_key_enabled = 0;--
2.34.1Reviewed-by Will Cohen <[email protected]>
Re: [PATCH v7 20/22] ui/cocoa: capture all keys and combos when mouse is grabbed
In reply to Philippe Mathieu-DaudéQuoted 205 lines · Philippe Mathieu-Daudé
From: Gustavo Noronha Silva <[email protected]>
Applications such as Gnome may use Alt-Tab and Super-Tab for different purposes, some use Ctrl-arrows so we want to allow qemu to handle everything when it captures the mouse/keyboard.
However, Mac OS handles some combos like Command-Tab and Ctrl-arrows at an earlier part of the event handling chain, not letting qemu see it.
We add a global Event Tap that allows qemu to see all events when the mouse is grabbed. Note that this requires additional permissions.
See:
https://developer.apple.com/documentation/coregraphics/1454426-cgeventtapcreate?language=objc#discussion
https://support.apple.com/en-in/guide/mac-help/mh32356/macSigned-off-by Gustavo Noronha Silva <[email protected]>Message-Id: <[email protected]>
Signed-off-by Akihiko Odaki <[email protected]>Message-Id: <[email protected]>
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- qapi/ui.json | 8 +++++- qemu-options.hx | 3 +++ ui/cocoa.m | 65 ++++++++++++++++++++++++++++++++++++++++++++++++- 3 files changed, 74 insertions(+), 2 deletions(-)qapi/ui.json +4
-0@@ -1270,11 +1270,17 @@ # host without sending this key to the guest when # "off". Defaults to "on" # +# @full-grab: Capture all key presses, including system combos. This +# requires accessibility permissions, since it performs +# a global grab on key events. (default: off) +# Seehttps://support.apple.com/en-in/guide/mac-help/mh32356/mac
qapi/ui.json +3
-1+# # Since: 7.0 ## { 'struct': 'DisplayCocoa', 'data': { - '*left-command-key': 'bool' + '*left-command-key': 'bool', + '*full-grab': 'bool' } } ##qemu-options.hx +3
-0@@ -1916,6 +1916,9 @@ DEF("display", HAS_ARG, QEMU_OPTION_display, #if defined(CONFIG_CURSES) "-display curses[,charset=<encoding>]\n" #endif +#if defined(CONFIG_COCOA) + "-display cocoa[,full_grab=on|off]\n" +#endif #if defined(CONFIG_OPENGL) "-display egl-headless[,rendernode=<file>]\n" #endifui/cocoa.m +15
-0@@ -309,11 +309,13 @@ static void handleAnyDeviceErrors(Error * err) BOOL isMouseGrabbed; BOOL isFullscreen; BOOL isAbsoluteEnabled; + CFMachPortRef eventsTap; } - (void) switchSurface:(pixman_image_t *)image; - (void) grabMouse; - (void) ungrabMouse; - (void) toggleFullScreen:(id)sender; +- (void) setFullGrab:(id)sender; - (void) handleMonitorInput:(NSEvent *)event; - (bool) handleEvent:(NSEvent *)event; - (bool) handleEventLocked:(NSEvent *)event;@@ -336,6 +338,19 @@ static void handleAnyDeviceErrors(Error * err) QemuCocoaView *cocoaView; +static CGEventRef handleTapEvent(CGEventTapProxy proxy, CGEventType type, CGEventRef cgEvent, void *userInfo) +{ + QemuCocoaView *cocoaView = userInfo; + NSEvent *event = [NSEvent eventWithCGEvent:cgEvent]; + if ([cocoaView isMouseGrabbed] && [cocoaView handleEvent:event]) { + COCOA_DEBUG("Global events tap: qemu handled the event, capturing!\n"); + return NULL; + } + COCOA_DEBUG("Global events tap: qemu did not handle the event, letting it through...\n"); + + return cgEvent; +} +@implementation QemuCocoaView
- (id)initWithFrame:(NSRect)frameRect
{ui/cocoa.m +17
-0@@ -361,6 +376,11 @@ QemuCocoaView *cocoaView; } qkbd_state_free(kbd); + + if (eventsTap) { + CFRelease(eventsTap); + } + [super dealloc]; }@@ -655,6 +675,36 @@ QemuCocoaView *cocoaView; } } +- (void) setFullGrab:(id)sender +{ + COCOA_DEBUG("QemuCocoaView: setFullGrab\n"); + + CGEventMask mask = CGEventMaskBit(kCGEventKeyDown) | CGEventMaskBit(kCGEventKeyUp) | CGEventMaskBit(kCGEventFlagsChanged); + eventsTap = CGEventTapCreate(kCGHIDEventTap, kCGHeadInsertEventTap, kCGEventTapOptionDefault, + mask, handleTapEvent, self); + if (!eventsTap) { + warn_report("Could not create event tap, system key combos will not be captured.\n"); + return; + } else { + COCOA_DEBUG("Global events tap created! Will capture system keycombos.\n");
ui/cocoa.m +5
-0+ } + + CFRunLoopRef runLoop = CFRunLoopGetCurrent(); + if (!runLoop) { + warn_report("Could not obtain current CF RunLoop, system keycombos will not be captured.\n");
ui/cocoa.m +4
-0+ return; + } + + CFRunLoopSourceRef tapEventsSrc =
CFMachPortCreateRunLoopSource(kCFAllocatorDefault, eventsTap, 0);ui/cocoa.m +2
-0+ if (!tapEventsSrc ) { + warn_report("Could not obtain current CF RunLoop, system keycombos will not be captured.\n");
ui/cocoa.m +21
-1+ return; + } + + CFRunLoopAddSource(runLoop, tapEventsSrc, kCFRunLoopDefaultMode); + CFRelease(tapEventsSrc); +} + - (void) toggleKey: (int)keycode { qkbd_state_key_event(kbd, keycode, !qkbd_state_key_get(kbd, keycode)); }@@ -1284,6 +1334,13 @@ QemuCocoaView *cocoaView; [cocoaView toggleFullScreen:sender]; } +- (void) setFullGrab:(id)sender +{ + COCOA_DEBUG("QemuCocoaAppController: setFullGrab\n"); + + [cocoaView setFullGrab:sender]; +} + /* Tries to find then open the specified filename */ - (void) openDocumentation: (NSString *) filename {@@ -2060,11 +2117,17 @@ static void cocoa_display_init(DisplayState *ds, DisplayOptions *opts) qemu_sem_wait(&app_started_sem); COCOA_DEBUG("cocoa_display_init: app start completed\n"); + QemuCocoaAppController *controller = (QemuCocoaAppController *)[[NSApplication sharedApplication] delegate]; /* if fullscreen mode is to be used */ if (opts->has_full_screen && opts->full_screen) { dispatch_async(dispatch_get_main_queue(), ^{ [NSApp activateIgnoringOtherApps: YES]; - [(QemuCocoaAppController *)[[NSApplication sharedApplication] delegate] toggleFullScreen: nil]; + [controller toggleFullScreen: nil]; + }); + } + if (opts->u.cocoa.has_full_grab && opts->u.cocoa.full_grab) { + dispatch_async(dispatch_get_main_queue(), ^{ + [controller setFullGrab: nil];}); }--
2.34.1Reviewed-by Will Cohen <[email protected]>
Re: [PATCH v7 06/22] hvf: Fix OOB write in RDTSCP instruction decode
In reply to Philippe Mathieu-DaudéHi Paolo,
I forgot to Cc you. Could you Ack this patch?
Quoted 106 lines · Philippe Mathieu-Daudé
From: Cameron Esfahani <[email protected]>
A guest could craft a specific stream of instructions that will have QEMU write
0xF9to inappropriate locations in memory. Add additional asserts to check for this. Generate a #UD if there are more than 14 prefix bytes.Found by Julian Stecklina <[email protected]>
Signed-off-by Cameron Esfahani <[email protected]>Message-Id: <[email protected]>
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- target/i386/hvf/x86_decode.c | 12 ++++++++++-- target/i386/hvf/x86hvf.c | 8 ++++++++ target/i386/hvf/x86hvf.h | 1 + 3 files changed, 19 insertions(+), 2 deletions(-)target/i386/hvf/x86_decode.c +10
-2@@ -24,8 +24,10 @@#include "vmx.h" #include "x86_mmu.h" #include "x86_descr.h" +#include "x86hvf.h" #define OPCODE_ESCAPE 0xf +#define X86_MAX_INSN_PREFIX_LENGTH (14)static void decode_invalid (CPUX86State *env,struct x86_decode *decode) {@@ -541,7 +543,8 @@ static void decode_lidtgroup(CPUX86State *env, struct x86_decode *decode) }; decode->cmd = group[decode->modrm.reg];if (0xf9 == decode->modrm.modrm) { - decode->opcode[decode->len++] = decode->modrm.modrm; +VM_PANIC_ON (decode->opcode_len >=sizeof (decode->opcode)); + decode->opcode[decode->opcode_len++] = decode->modrm.modrm; decode->cmd =X86_DECODE_CMD_RDTSCP ; } }@@ -1847,7 +1850,8 @@ void calc_modrm_operand(CPUX86State *env, struct x86_decode *decode,static void decode_prefix (CPUX86State *env,struct x86_decode *decode) { -while (1 ) { +/* At most X86_MAX_INSN_PREFIX_LENGTH prefix bytes. */ +for (int i =0 ; i <X86_MAX_INSN_PREFIX_LENGTH ; i++) {/* * REX prefix must come after legacy prefixes. * REX before legacy is ignored. @@ -1892,6 +1896,8 @@ static void decode_prefix(CPUX86State *env, struct x86_decode *decode)return ; } } +/* Too many prefixes! Generate #UD. */ +hvf_inject_ud (env); }void set_addressing_size (CPUX86State *env,struct x86_decode *decode)@@ -2090,11 +2096,13 @@ static void decode_opcodes(CPUX86State *env, struct x86_decode *decode)uint8_t opcode; opcode =decode_byte (env, decode); +VM_PANIC_ON (decode->opcode_len >=sizeof (decode->opcode)); decode->opcode[decode->opcode_len++] = opcode;if (opcode !=OPCODE_ESCAPE ) {decode_opcode_1 (env, decode, opcode); }else { opcode =decode_byte (env, decode); +VM_PANIC_ON (decode->opcode_len >=sizeof (decode->opcode)); decode->opcode[decode->opcode_len++] = opcode;decode_opcode_2 (env, decode, opcode); }target/i386/hvf/x86hvf.c +8
-0@@ -423,6 +423,14 @@ bool hvf_inject_interrupts(CPUState *cpu_state) & (CPU_INTERRUPT_INIT | CPU_INTERRUPT_TPR)); } +void hvf_inject_ud (CPUX86State *env) +{ + env->exception_nr =EXCP06_ILLOP ; + env->exception_injected =1 ; + env->has_error_code = false; + env->error_code =0 ; +} +int hvf_process_events (CPUState *cpu_state) {X86CPU *cpu =X86_CPU (cpu_state);target/i386/hvf/x86hvf.h +1
-0@@ -22,6 +22,7 @@int hvf_process_events (CPUState *);bool hvf_inject_interrupts (CPUState *); +void hvf_inject_ud (CPUX86State *);void hvf_set_segment (struct CPUState *cpu,struct vmx_segment *vmx_seg,SegmentCache *qseg,bool is_tr);void hvf_get_segment (SegmentCache *qseg,struct vmx_segment *vmx_seg);
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
In reply to Philippe Mathieu-DaudéHi Alex, Thomas, Daniel,
Could you ack this patch?
Quoted 80 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Add support for macOS 12 build on Cirrus-CI, similarly to commit
0e103a65ba1("gitlab: support for ... macOS 11 via cirrus-run"), but with the following differences:
- Enable modules (configure --enable-modules)
- Do not run softfloat3 tests (make check-softfloat)
- Run Aarch64 qtests instead ofx86_64onesGenerate the vars file by calling 'make lcitool-refresh'.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- .gitlab-ci.d/cirrus.yml | 16 ++++++++++++++++ .gitlab-ci.d/cirrus/macos-12.vars | 16 ++++++++++++++++ tests/lcitool/refresh | 1 + 3 files changed, 33 insertions(+)create mode 100644 .gitlab-ci.d/cirrus/macos-12.vars
.gitlab-ci.d/cirrus.yml +16
-0@@ -87,6 +87,22 @@ x64-macos-11-base-build: PKG_CONFIG_PATH: /usr/local/opt/curl/lib/pkgconfig:/usr/local/opt/ncurses/lib/pkgconfig:/usr/local/opt/readline/lib/pkgconfig TEST_TARGETS: check-unit check-block check-qapi-schema check-softfloat check-qtest-x86_64 +x64-macos-12-base-build: + extends: .cirrus_build_job + variables: + NAME: macos-12 + CIRRUS_VM_INSTANCE_TYPE: osx_instance + CIRRUS_VM_IMAGE_SELECTOR: image + CIRRUS_VM_IMAGE_NAME: monterey-base + CIRRUS_VM_CPUS: 12 + CIRRUS_VM_RAM: 24G + UPDATE_COMMAND: brew update + INSTALL_COMMAND: brew install + PATH_EXTRA: /usr/local/opt/ccache/libexec:/usr/local/opt/gettext/bin + PKG_CONFIG_PATH: /usr/local/opt/curl/lib/pkgconfig:/usr/local/opt/ncurses/lib/pkgconfig:/usr/local/opt/readline/lib/pkgconfig + CONFIGURE_ARGS: --enable-modules + TEST_TARGETS: check-unit check-block check-qapi-schema check-qtest-aarch64 + # The following jobs run VM-based tests via KVM on a Linux-based Cirrus-CI job .cirrus_kvm_job:.gitlab-ci.d/cirrus/macos-12.vars new file +16
-0@@ -0,0 +1,16 @@ +# THIS FILE WAS AUTO-GENERATED +# +# $ lcitool variables macos-12 qemu +# +# https://gitlab.com/libvirt/libvirt-ci + +CCACHE='/usr/local/bin/ccache' +CPAN_PKGS='Test::Harness' +CROSS_PKGS='' +MAKE='/usr/local/bin/gmake' +NINJA='/usr/local/bin/ninja' +PACKAGING_COMMAND='brew' +PIP3='/usr/local/bin/pip3' +PKGS='bash bc bzip2 capstone ccache cpanminus ctags curl dbus diffutils dtc gcovr gettext git glib gnu-sed gnutls gtk+3 jemalloc jpeg-turbo libepoxy libffi libgcrypt libiscsi libnfs libpng libslirp libssh libtasn1 libusb llvm lzo make meson ncurses nettle ninja perl pixman pkg-config python3 rpm2cpio sdl2 sdl2_image snappy sparse spice-protocol tesseract texinfo usbredir vde vte3 zlib zstd' +PYPI_PKGS='PyYAML numpy pillow sphinx sphinx-rtd-theme virtualenv' +PYTHON='/usr/local/bin/python3'tests/lcitool/refresh +1
-0@@ -105,6 +105,7 @@ try: generate_cirrus("freebsd-12") generate_cirrus("freebsd-13") generate_cirrus("macos-11") + generate_cirrus("macos-12") sys.exit(0) except Exception as ex:
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
In reply to Philippe Mathieu-DaudéQuoted 11 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Add support for macOS 12 build on Cirrus-CI, similarly to commit
0e103a65ba1("gitlab: support for ... macOS 11 via cirrus-run"), but with the following differences:
- Enable modules (configure --enable-modules)
- Do not run softfloat3 tests (make check-softfloat)
- Run Aarch64 qtests instead ofx86_64onesGenerate the vars file by calling 'make lcitool-refresh'.
Quoted 7 lines
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- .gitlab-ci.d/cirrus.yml | 16 ++++++++++++++++ .gitlab-ci.d/cirrus/macos-12.vars | 16 ++++++++++++++++ tests/lcitool/refresh | 1 + 3 files changed, 33 insertions(+) create mode 100644 .gitlab-ci.d/cirrus/macos-12.vars
Regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH v7 02/22] tests/fp/berkeley-testfloat-3: Ignore ignored #pragma directives
In reply to Philippe Mathieu-DaudéQuoted 14 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Since we already use -Wno-unknown-pragmas, we can also use -Wno-ignored-pragmas. This silences hundred of warnings using clang 13 on macOS Monterey:
[409/771] Compiling C object
tests/fp/libtestfloat.a.p/berkeley-testfloat-3_source_test_az_f128_rx.c.o../tests/fp/berkeley-testfloat-3/source/test_az_f128_rx.c:49:14: warning: '#pragmaFENV_ACCESS' is not supported on this target - ignored [-Wignored-pragmas] #pragma STDCFENV_ACCESSON^1 warning generated.
Having:
Quoted 7 lines
$ cc -v Apple clang version 13.0.0 (clang-1300.0.29.30)Reported-by Roman Bolshakov <[email protected]>Reviewed-by Akihiko Odaki <[email protected]>Tested-by Akihiko Odaki <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>
Signature · 1 line
Alex Bennée
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
In reply to Philippe Mathieu-DaudéQuoted 3 lines · Philippe Mathieu-Daudé
Hi Alex, Thomas, Daniel,
Could you ack this patch?
Basically fine for me, but can we really run additional cirrus-ci jobs by default? IIRC the parallel execution of those were quite limited for the free tier, so did you look close that we don't run into additional timeouts yet, due to delayed cirrus-ci jobs?
Thomas
Quoted 19 lines · Philippe Mathieu-Daudé
Quoted 18 lines
From: Philippe Mathieu-Daudé <[email protected]>
Add support for macOS 12 build on Cirrus-CI, similarly to commit
0e103a65ba1("gitlab: support for ... macOS 11 via cirrus-run"), but with the following differences:
- Enable modules (configure --enable-modules)
- Do not run softfloat3 tests (make check-softfloat)
- Run Aarch64 qtests instead ofx86_64onesGenerate the vars file by calling 'make lcitool-refresh'.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>---
.gitlab-ci.d/cirrus.yml | 16 ++++++++++++++++ .gitlab-ci.d/cirrus/macos-12.vars | 16 ++++++++++++++++ tests/lcitool/refresh | 1 +3 files changed, 33 insertions(+)create mode 100644 .gitlab-ci.d/cirrus/macos-12.vars
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
Quoted 9 lines · Thomas Huth
Quoted 3 lines · Philippe Mathieu-Daudé
Hi Alex, Thomas, Daniel,
Could you ack this patch?
Basically fine for me, but can we really run additional cirrus-ci jobs by default? IIRC the parallel execution of those were quite limited for the free tier, so did you look close that we don't run into additional timeouts yet, due to delayed cirrus-ci jobs?
You can run 2 jobs in parallel in Cirrus. Beyond that they get queued/serialized
We have a 1 hour job timeout.
We have to expect jobs will sometimes run slower than normal.
IOW if we have a job on Cirrus taking 30 minutes normally, we expect it will sometimes take 45 minutes.
All this means that if we want Cirrus to be reliable and not time out, we can really only have 2 jobs by default, unless we can get the job execution time down to around 20 minutes to allow for serialization.
We used to have terrible problems with cirrus timeouts when we were running 4 jobs concurrently (2 on staging and 2 on master). We addressed that in 9968de0a4a5470bd7b98dcd2fae5d5269908f16b by disabling the jobs on master.
IOW, we really need to choose 1 macOS job and 1 FreeBSD job and others need to be marked manual.
Regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
Quoted 22 lines · Daniel P. Berrangé
Quoted 4 lines · Thomas Huth
Basically fine for me, but can we really run additional cirrus-ci jobs by default? IIRC the parallel execution of those were quite limited for the free tier, so did you look close that we don't run into additional timeouts yet, due to delayed cirrus-ci jobs?
You can run 2 jobs in parallel in Cirrus. Beyond that they get queued/serialized
We have a 1 hour job timeout.
We have to expect jobs will sometimes run slower than normal.
IOW if we have a job on Cirrus taking 30 minutes normally, we expect it will sometimes take 45 minutes.
All this means that if we want Cirrus to be reliable and not time out, we can really only have 2 jobs by default, unless we can get the job execution time down to around 20 minutes to allow for serialization.
Quoted 7 lines
We used to have terrible problems with cirrus timeouts when we were running 4 jobs concurrently (2 on staging and 2 on master). We addressed that in
9968de0a4a5470bd7b98dcd2fae5d5269908f16bby disabling the jobs on master.IOW, we really need to choose 1 macOS job and 1 FreeBSD job and others need to be marked manual.
Not sure which job to choose yet. Per the first google hits we still want to cover Catalina first:
https://www.statista.com/statistics/944559/worldwide-macos-version-market-share/
Would it be beneficial to have a 1 per OS job during PR, and other jobs run nightly (once a day, not per PR)?
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
Quoted 2 lines · Philippe Mathieu-Daudé
Would it be beneficial to have a 1 per OS job during PR, and other jobs run nightly (once a day, not per PR)?
Is there a way to trigger nightly runs in gitlab?
Thomas
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
Quoted 6 lines · Thomas Huth
Quoted 2 lines · Philippe Mathieu-Daudé
Would it be beneficial to have a 1 per OS job during PR, and other jobs run nightly (once a day, not per PR)?
Is there a way to trigger nightly runs in gitlab?
Yes, in the repo CI settings create a scheduled job, using cron syntax. Relies on the job "rules" in the config file not excluding scheduled execution of course.
Regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
In reply to Philippe Mathieu-DaudéQuoted 28 lines · Philippe Mathieu-Daudé
Quoted 27 lines · Daniel P. Berrangé
Quoted 4 lines · Thomas Huth
Basically fine for me, but can we really run additional cirrus-ci jobs by default? IIRC the parallel execution of those were quite limited for the free tier, so did you look close that we don't run into additional timeouts yet, due to delayed cirrus-ci jobs?
You can run 2 jobs in parallel in Cirrus. Beyond that they get queued/serialized
We have a 1 hour job timeout.
We have to expect jobs will sometimes run slower than normal.
IOW if we have a job on Cirrus taking 30 minutes normally, we expect it will sometimes take 45 minutes.
All this means that if we want Cirrus to be reliable and not time out, we can really only have 2 jobs by default, unless we can get the job execution time down to around 20 minutes to allow for serialization.
We used to have terrible problems with cirrus timeouts when we were running 4 jobs concurrently (2 on staging and 2 on master). We addressed that in
9968de0a4a5470bd7b98dcd2fae5d5269908f16bby disabling the jobs on master.
Quoted 6 lines
Quoted 2 lines
IOW, we really need to choose 1 macOS job and 1 FreeBSD job and others need to be marked manual.
Not sure which job to choose yet. Per the first google hits we still want to cover Catalina first:
https://www.statista.com/statistics/944559/worldwide-macos-version-market-share/
My general gut feeling is usually to prioritize testing older versions as they tend to be more widely used, and we want to avoid regresions on stuff that has been around the longest. Compatibility problems with new releases tend to get reported by users/maintainers and would not be regressions, but rather enhancements to support the new platform.
But it is not really an easy decision, as clearly we would prefer to test both old and new if we had the ability.
Having one as a nightly job is a reasonable compromise.
With regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
Quoted 5 lines · Daniel P. Berrangé
Quoted 3 lines · Philippe Mathieu-Daudé
Not sure which job to choose yet. Per the first google hits we still want to cover Catalina first:
https://www.statista.com/statistics/944559/worldwide-macos-version-market-share/
...that page says that the figures for Catalina are "incorrectly high", so not a very useful set of data for these purposes I think.
Quoted 5 lines
My general gut feeling is usually to prioritize testing older versions as they tend to be more widely used, and we want to avoid regresions on stuff that has been around the longest. Compatibility problems with new releases tend to get reported by users/maintainers and would not be regressions, but rather enhancements to support the new platform.
Note that technically speaking Catalina (10.15) has already fallen off the end of our support policy (which currently requires Big Sur (11) and Monterey (12) support, as the two most recent versions). My personal macos x86 laptop is still running Catalina, though, because Big Sur dropped support for that hardware :-/
That suggests that maybe Big Sur would be what we should go for now if we have to pick just one for the CI.
thanks
-- PMM
Re: [PATCH v7 22/22] gitlab-ci: Support macOS 12 via cirrus-run
Quoted 16 lines · Peter Maydell
Quoted 5 lines · Daniel P. Berrangé
Quoted 3 lines · Philippe Mathieu-Daudé
Not sure which job to choose yet. Per the first google hits we still want to cover Catalina first:
https://www.statista.com/statistics/944559/worldwide-macos-version-market-share/...that page says that the figures for Catalina are "incorrectly high", so not a very useful set of data for these purposes I think.
Quoted 5 lines
My general gut feeling is usually to prioritize testing older versions as they tend to be more widely used, and we want to avoid regresions on stuff that has been around the longest. Compatibility problems with new releases tend to get reported by users/maintainers and would not be regressions, but rather enhancements to support the new platform.
Quoted 7 lines
Note that technically speaking Catalina (10.15) has already fallen off the end of our support policy (which currently requires Big Sur (11) and Monterey (12) support, as the two most recent versions). My personal macos x86 laptop is still running Catalina, though, because Big Sur dropped support for that hardware :-/
That suggests that maybe Big Sur would be what we should go for now if we have to pick just one for the CI.
OK. Since I'm running Monterey (12) on my workstation, I'll make this job manual / nightly, and keep the Big Sur (11) on CI.
I'll let Daniel / Thomas / Alex decide which FreeBSD job to demote to manual.
Regards,
Phil.
Quoted 4 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Few patches to be able to build QEMU on macOS 12 (Monterey).
Quoted 10 lines
Missing review:
0006-hvf-Fix-OOB-write-in-RDTSCP-instruction-decode.patch 0013-osdep-Avoid-using-Clang-specific-__builtin_available.patch 0014-meson-Resolve-the-entitlement.sh-script-once-for-goo.patch 0015-meson-Log-QEMU_CXXFLAGS-content-in-summary.patch 0016-configure-Pass-filtered-QEMU_OBJCFLAGS-to-meson.patch 0017-ui-cocoa-Constify-qkeycode-translation-arrays.patch 0020-ui-cocoa-capture-all-keys-and-combos-when-mouse-is-g.patch 0021-ui-cocoa-add-option-to-swap-Option-and-Command.patch 0022-gitlab-ci-Support-macOS-12-via-cirrus-run.patch
Queuing reviewed patches; IOW all except:
0006-hvf-Fix-OOB-write-in-RDTSCP-instruction-decode.patch 0022-gitlab-ci-Support-macOS-12-via-cirrus-run.patch
Thanks,
Phil.
Re: [PATCH v7 06/22] hvf: Fix OOB write in RDTSCP instruction decode
In reply to Peter MaydellQuoted 11 lines · Peter Maydell
Quoted 4 lines · Philippe Mathieu-Daudé
Hi Paolo,
I forgot to Cc you. Could you Ack this patch?
I had review comments on the version of this patch in v5 which still need to be addressed:
Thanks for pointing at your comments, I totally missed them.
I'll let Cameron address them.
Regards,
Phil.
Re: [PATCH v7 06/22] hvf: Fix OOB write in RDTSCP instruction decode
In reply to Philippe Mathieu-DaudéQuoted 4 lines · Philippe Mathieu-Daudé
Hi Paolo,
I forgot to Cc you. Could you Ack this patch?
I had review comments on the version of this patch in v5 which still need to be addressed:
thanks
-- PMM
Quoted 15 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Few patches to be able to build QEMU on macOS 12 (Monterey).
Missing review:
0006-hvf-Fix-OOB-write-in-RDTSCP-instruction-decode.patch 0013-osdep-Avoid-using-Clang-specific-__builtin_available.patch 0014-meson-Resolve-the-entitlement.sh-script-once-for-goo.patch 0015-meson-Log-QEMU_CXXFLAGS-content-in-summary.patch 0016-configure-Pass-filtered-QEMU_OBJCFLAGS-to-meson.patch 0017-ui-cocoa-Constify-qkeycode-translation-arrays.patch 0020-ui-cocoa-capture-all-keys-and-combos-when-mouse-is-g.patch 0021-ui-cocoa-add-option-to-swap-Option-and-Command.patch 0022-gitlab-ci-Support-macOS-12-via-cirrus-run.patch
Quoted 5 lines
Since v6:
- Dropped merged patches
- Addressed Akihiko Odaki comments (squashed 2 patches, added R/T-b)
- Dropped 'configure: Disable out-of-line atomic operations on Aarch64'
- Add few macos patches on the list pending for 7.0 so tested by CI
Hi Philippe, I did not find v6 somehow,
and I was looking for patch:
"[PATCH v5 06/16] hvf: Enable RDTSCP support"
was it dropped / merged with something else? I do not see it in latest git, nor in this respin, maybe it is in your tree somewhere?
Thanks,
Claudio
Quoted 81 lines
Since v5:
- Fixed failed rebase between patches 10 and 16 (Akihiko)
- Include "ui/cocoa: Fix the leak ofqemu_console_get_label"Since v4:
- UseMAC_OS_X_VERSION_MIN_REQUIREDdefinition (Akihiko)
- Include patches from AkihikoSince v3:
- Fix --enable-modules
- Ignore #pragma on softfloat3 tests
- Addressed Akihiko Odaki comments
- Include Cameron Esfahani patchesSince v2:
- Addressed Akihiko Odaki comments:
. use__is_identifier(),
. remove cocoasetAllowedFileTypes()
- Addressed Daniel Berrangé comment:
. rebased on testing/next, update libvirt-ci/lcitoolAkihiko Odaki (2):
audio: Log context for audio bug
coreaudio: Always return 0 inhandle_voice_changeCameron Esfahani (2):
hvf: Use standard CR0 and CR4 register definitions
hvf: Fix OOB write in RDTSCP instruction decodeCarwyn Ellis (2):
ui/cocoa: add option to disable left-command forwarding to guest ui/cocoa: release mouse when user switches away from QEMU windowGustavo Noronha Silva (2):
ui/cocoa: capture all keys and combos when mouse is grabbed ui/cocoa: add option to swap Option and CommandPhilippe Mathieu-Daudé (14):
configure: Allow passing extra Objective C compiler flags tests/fp/berkeley-testfloat-3: Ignore ignored #pragma directives
hvf: Makehvf_get_segments()/hvf_put_segments()local
hvf: Remove deprecatedhv_vcpu_flush()callsblock/file-posix: Remove a deprecation warning on macOS 12 audio/coreaudio: Remove a deprecation warning on macOS 12 audio/dbus: Fix building with modules on macOS
audio: Rename coreaudio extension to use Objective-C compiler
osdep: Avoid using Clang-specific__builtin_available()
meson: Resolve theentitlement.shscript once for good
meson: LogQEMU_CXXFLAGScontent in summary
configure: Pass filteredQEMU_OBJCFLAGSto meson ui/cocoa: Constify qkeycode translation arrays
gitlab-ci: Support macOS 12 via cirrus-run.gitlab-ci.d/cirrus.yml | 16 ++++ .gitlab-ci.d/cirrus/macos-12.vars | 16 ++++ audio/audio.c | 25 +++--- audio/audio_template.h | 27 +++---- audio/{coreaudio.c => coreaudio.m} | 23 +++--- audio/meson.build | 4 +- block/file-posix.c | 14 +++- configure | 31 ++++++++ include/qemu/osdep.h | 10 +-- meson.build | 17 +++- qapi/ui.json | 29 +++++++ qemu-options.hx | 15 ++++ target/i386/hvf/vmx.h | 19 +++-- target/i386/hvf/x86.c | 6 +- target/i386/hvf/x86.h | 34 -------- target/i386/hvf/x86_decode.c | 12 ++- target/i386/hvf/x86_mmu.c | 2 +- target/i386/hvf/x86_task.c | 4 +- target/i386/hvf/x86hvf.c | 14 +++- target/i386/hvf/x86hvf.h | 3 +- tests/fp/meson.build | 5 ++ tests/lcitool/refresh | 1 + ui/cocoa.m | 122 ++++++++++++++++++++++++++--- 23 files changed, 327 insertions(+), 122 deletions(-) create mode 100644 .gitlab-ci.d/cirrus/macos-12.vars rename audio/{coreaudio.c => coreaudio.m} (97%)
Quoted 23 lines · Claudio Fontana
Quoted 15 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Few patches to be able to build QEMU on macOS 12 (Monterey).
Missing review:
0006-hvf-Fix-OOB-write-in-RDTSCP-instruction-decode.patch 0013-osdep-Avoid-using-Clang-specific-__builtin_available.patch 0014-meson-Resolve-the-entitlement.sh-script-once-for-goo.patch 0015-meson-Log-QEMU_CXXFLAGS-content-in-summary.patch 0016-configure-Pass-filtered-QEMU_OBJCFLAGS-to-meson.patch 0017-ui-cocoa-Constify-qkeycode-translation-arrays.patch 0020-ui-cocoa-capture-all-keys-and-combos-when-mouse-is-g.patch 0021-ui-cocoa-add-option-to-swap-Option-and-Command.patch 0022-gitlab-ci-Support-macOS-12-via-cirrus-run.patchQuoted 5 lines
Since v6:
- Dropped merged patches
- Addressed Akihiko Odaki comments (squashed 2 patches, added R/T-b)
- Dropped 'configure: Disable out-of-line atomic operations on Aarch64'
- Add few macos patches on the list pending for 7.0 so tested by CI
Quoted 8 lines
Hi Philippe, I did not find v6 somehow,
and I was looking for patch:
"[PATCH v5 06/16] hvf: Enable RDTSCP support"
was it dropped / merged with something else? I do not see it in latest git, nor in this respin, maybe it is in your tree somewhere?
The patch stayed unreviewed during 2 months, so I dropped it.
Now it got at least a Tested-by tag from Silvio, I'll include it in the next PR.
Regards,
Phil.
Quoted 33 lines · Philippe Mathieu-Daudé
Quoted 23 lines · Claudio Fontana
Quoted 15 lines · Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <[email protected]>
Few patches to be able to build QEMU on macOS 12 (Monterey).
Missing review:
0006-hvf-Fix-OOB-write-in-RDTSCP-instruction-decode.patch 0013-osdep-Avoid-using-Clang-specific-__builtin_available.patch 0014-meson-Resolve-the-entitlement.sh-script-once-for-goo.patch 0015-meson-Log-QEMU_CXXFLAGS-content-in-summary.patch 0016-configure-Pass-filtered-QEMU_OBJCFLAGS-to-meson.patch 0017-ui-cocoa-Constify-qkeycode-translation-arrays.patch 0020-ui-cocoa-capture-all-keys-and-combos-when-mouse-is-g.patch 0021-ui-cocoa-add-option-to-swap-Option-and-Command.patch 0022-gitlab-ci-Support-macOS-12-via-cirrus-run.patchQuoted 5 lines
Since v6:
- Dropped merged patches
- Addressed Akihiko Odaki comments (squashed 2 patches, added R/T-b)
- Dropped 'configure: Disable out-of-line atomic operations on Aarch64'
- Add few macos patches on the list pending for 7.0 so tested by CIQuoted 8 lines
Hi Philippe, I did not find v6 somehow,
and I was looking for patch:
"[PATCH v5 06/16] hvf: Enable RDTSCP support"
was it dropped / merged with something else? I do not see it in latest git, nor in this respin, maybe it is in your tree somewhere?
Quoted 9 lines
The patch stayed unreviewed during 2 months, so I dropped it.
Now it got at least a Tested-by tag from Silvio, I'll include it in the next PR.
Regards,
Phil.
Awesome, thanks for the update!
Claudio