[Qemu-devel] [PATCH v3 0/8] Poison some more target-specific defines

[Qemu-devel] [PATCH v3 0/8] Poison some more target-specific defines

[Thomas Huth]

This series marks some more #defines as poisoned, which are
target-specific (declared in config-target.h) and thus must
not be used in common code.

v3:
- Mark some more TARGET_xxx and CONFIG_xxx defines as poisoned in
  the first two patches (as requested by Richard)
- Added the final patch to move the edu device to common-obj-y

v2:
- First two patches are the same as in v1
- Reworked the CONFIG_KVM patches according to Paolo's review feedback
- Added two new patches to finally poison CONFIG_SOFTMMU, too
- Added a final patch to move bootdevice.o to common-obj now
  (based on an earlier patch where I also tried to move numa.o and
  balloon.o, too - but these files are indirectly target-dependent as
  I now know, so they can't be moved)

Thomas Huth (8):
  include/exec/poison: Add missing TARGET defines
  include/exec/poison: Mark some CONFIG defines as poisoned, too
  Move CONFIG_KVM related definitions to kvm_i386.h
  include/exec/poison: Mark CONFIG_KVM as poisoned, too
  cpu: Introduce a wrapper for tlb_flush() that can be used in common
    code
  include/exec/poison: Mark CONFIG_SOFTMMU as poisoned
  Makefile: Move bootdevice.o to common-obj-y
  hw/misc/edu: Compile the edu device as common object

 Makefile.objs             |  2 +-
 Makefile.target           |  2 +-
 bootdevice.c              |  2 +-
 hw/acpi/ich9.c            |  1 -
 hw/i386/pc_q35.c          |  1 +
 hw/misc/Makefile.objs     |  2 +-
 include/exec/cpu-common.h |  2 ++
 include/exec/poison.h     | 39 +++++++++++++++++++++++++++++++++++++++
 include/hw/i386/pc.h      | 13 -------------
 include/qom/cpu.h         |  8 ++++++++
 include/sysemu/kvm.h      | 31 ++++++++++++-------------------
 qom/cpu.c                 |  5 ++---
 target/i386/kvm_i386.h    | 23 +++++++++++++++++++++++
 translate-all.c           |  8 ++++++++
 14 files changed, 99 insertions(+), 40 deletions(-)

-- 
1.8.3.1

[Qemu-devel] [PATCH v3 1/8] include/exec/poison: Add missing TARGET defines

[Thomas Huth]

Since we've got some new CPU targets in QEMU during the last months
and years, we've got some new TARGET_xxx defines now which should
be marked as poisoned for common code.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 include/exec/poison.h | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/include/exec/poison.h b/include/exec/poison.h
index 3ca7929..859bd56 100644
--- a/include/exec/poison.h
+++ b/include/exec/poison.h
@@ -7,22 +7,38 @@
 
 #pragma GCC poison TARGET_I386
 #pragma GCC poison TARGET_X86_64
+#pragma GCC poison TARGET_AARCH64
 #pragma GCC poison TARGET_ALPHA
 #pragma GCC poison TARGET_ARM
 #pragma GCC poison TARGET_CRIS
+#pragma GCC poison TARGET_HPPA
 #pragma GCC poison TARGET_LM32
 #pragma GCC poison TARGET_M68K
+#pragma GCC poison TARGET_MICROBLAZE
 #pragma GCC poison TARGET_MIPS
+#pragma GCC poison TARGET_ABI_MIPSN32
+#pragma GCC poison TARGET_ABI_MIPSO32
 #pragma GCC poison TARGET_MIPS64
+#pragma GCC poison TARGET_ABI_MIPSN64
+#pragma GCC poison TARGET_MOXIE
+#pragma GCC poison TARGET_NIOS2
 #pragma GCC poison TARGET_OPENRISC
 #pragma GCC poison TARGET_PPC
 #pragma GCC poison TARGET_PPCEMB
 #pragma GCC poison TARGET_PPC64
 #pragma GCC poison TARGET_ABI32
+#pragma GCC poison TARGET_S390X
 #pragma GCC poison TARGET_SH4
 #pragma GCC poison TARGET_SPARC
 #pragma GCC poison TARGET_SPARC64
+#pragma GCC poison TARGET_TILEGX
+#pragma GCC poison TARGET_TRICORE
+#pragma GCC poison TARGET_UNICORE32
+#pragma GCC poison TARGET_XTENSA
 
+#pragma GCC poison TARGET_HAS_BFLT
+#pragma GCC poison TARGET_NAME
+#pragma GCC poison TARGET_SUPPORTS_MTTCG
 #pragma GCC poison TARGET_WORDS_BIGENDIAN
 #pragma GCC poison BSWAP_NEEDED
 
-- 
1.8.3.1

[Qemu-devel] [PATCH v3 2/8] include/exec/poison: Mark some CONFIG defines as poisoned, too

[Thomas Huth]

These are defined in config-target.h and thus should never be
used in common code.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 include/exec/poison.h | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/include/exec/poison.h b/include/exec/poison.h
index 859bd56..7a025b2 100644
--- a/include/exec/poison.h
+++ b/include/exec/poison.h
@@ -66,5 +66,26 @@
 #pragma GCC poison CPU_INTERRUPT_TGT_INT_1
 #pragma GCC poison CPU_INTERRUPT_TGT_INT_2
 
+#pragma GCC poison CONFIG_ALPHA_DIS
+#pragma GCC poison CONFIG_ARM_A64_DIS
+#pragma GCC poison CONFIG_ARM_DIS
+#pragma GCC poison CONFIG_CRIS_DIS
+#pragma GCC poison CONFIG_HPPA_DIS
+#pragma GCC poison CONFIG_I386_DIS
+#pragma GCC poison CONFIG_LM32_DIS
+#pragma GCC poison CONFIG_M68K_DIS
+#pragma GCC poison CONFIG_MICROBLAZE_DIS
+#pragma GCC poison CONFIG_MIPS_DIS
+#pragma GCC poison CONFIG_MOXIE_DIS
+#pragma GCC poison CONFIG_NIOS2_DIS
+#pragma GCC poison CONFIG_PPC_DIS
+#pragma GCC poison CONFIG_S390_DIS
+#pragma GCC poison CONFIG_SH4_DIS
+#pragma GCC poison CONFIG_SPARC_DIS
+#pragma GCC poison CONFIG_XTENSA_DIS
+
+#pragma GCC poison CONFIG_LINUX_USER
+#pragma GCC poison CONFIG_VHOST_NET
+
 #endif
 #endif
-- 
1.8.3.1

[Qemu-devel] [PATCH v3 3/8] Move CONFIG_KVM related definitions to kvm_i386.h

[Thomas Huth]

pc.h and sysemu/kvm.h are also included from common code (where
CONFIG_KVM is not available), so the #defines that depend on CONFIG_KVM
should not be declared here to avoid that anybody is using them in a
wrong way. Since we're also going to poison CONFIG_KVM for common code,
let's move them to kvm_i386.h instead. Most of the dummy definitions
from sysemu/kvm.h are also unused since the code that uses them is
only compiled for CONFIG_KVM (e.g. target/i386/kvm.c), so the unused
defines are also simply dropped here instead of being moved.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 hw/i386/pc_q35.c       |  1 +
 include/hw/i386/pc.h   | 13 -------------
 include/sysemu/kvm.h   | 15 ---------------
 target/i386/kvm_i386.h | 23 +++++++++++++++++++++++
 4 files changed, 24 insertions(+), 28 deletions(-)

diff --git a/hw/i386/pc_q35.c b/hw/i386/pc_q35.c
index 1523ef3..8f696b7 100644
--- a/hw/i386/pc_q35.c
+++ b/hw/i386/pc_q35.c
@@ -36,6 +36,7 @@
 #include "hw/timer/mc146818rtc.h"
 #include "hw/xen/xen.h"
 #include "sysemu/kvm.h"
+#include "kvm_i386.h"
 #include "hw/kvm/clock.h"
 #include "hw/pci-host/q35.h"
 #include "exec/address-spaces.h"
diff --git a/include/hw/i386/pc.h b/include/hw/i386/pc.h
index d071c9c..a31f7aa 100644
--- a/include/hw/i386/pc.h
+++ b/include/hw/i386/pc.h
@@ -20,19 +20,6 @@
 
 #define HPET_INTCAP "hpet-intcap"
 
-#ifdef CONFIG_KVM
-#define kvm_pit_in_kernel() \
-    (kvm_irqchip_in_kernel() && !kvm_irqchip_is_split())
-#define kvm_pic_in_kernel()  \
-    (kvm_irqchip_in_kernel() && !kvm_irqchip_is_split())
-#define kvm_ioapic_in_kernel() \
-    (kvm_irqchip_in_kernel() && !kvm_irqchip_is_split())
-#else
-#define kvm_pit_in_kernel()      0
-#define kvm_pic_in_kernel()      0
-#define kvm_ioapic_in_kernel()   0
-#endif
-
 /**
  * PCMachineState:
  * @acpi_dev: link to ACPI PM device that performs ACPI hotplug handling
diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h
index 1e91613..ca40b6e 100644
--- a/include/sysemu/kvm.h
+++ b/include/sysemu/kvm.h
@@ -22,21 +22,6 @@
 #ifdef CONFIG_KVM
 #include <linux/kvm.h>
 #include <linux/kvm_para.h>
-#else
-/* These constants must never be used at runtime if kvm_enabled() is false.
- * They exist so we don't need #ifdefs around KVM-specific code that already
- * checks kvm_enabled() properly.
- */
-#define KVM_CPUID_SIGNATURE      0
-#define KVM_CPUID_FEATURES       0
-#define KVM_FEATURE_CLOCKSOURCE  0
-#define KVM_FEATURE_NOP_IO_DELAY 0
-#define KVM_FEATURE_MMU_OP       0
-#define KVM_FEATURE_CLOCKSOURCE2 0
-#define KVM_FEATURE_ASYNC_PF     0
-#define KVM_FEATURE_STEAL_TIME   0
-#define KVM_FEATURE_PV_EOI       0
-#define KVM_FEATURE_CLOCKSOURCE_STABLE_BIT 0
 #endif
 
 extern bool kvm_allowed;
diff --git a/target/i386/kvm_i386.h b/target/i386/kvm_i386.h
index bfce427..1de9876 100644
--- a/target/i386/kvm_i386.h
+++ b/target/i386/kvm_i386.h
@@ -15,6 +15,29 @@
 
 #define kvm_apic_in_kernel() (kvm_irqchip_in_kernel())
 
+#ifdef CONFIG_KVM
+
+#define kvm_pit_in_kernel() \
+    (kvm_irqchip_in_kernel() && !kvm_irqchip_is_split())
+#define kvm_pic_in_kernel()  \
+    (kvm_irqchip_in_kernel() && !kvm_irqchip_is_split())
+#define kvm_ioapic_in_kernel() \
+    (kvm_irqchip_in_kernel() && !kvm_irqchip_is_split())
+
+#else
+
+#define kvm_pit_in_kernel()      0
+#define kvm_pic_in_kernel()      0
+#define kvm_ioapic_in_kernel()   0
+
+/* These constants must never be used at runtime if kvm_enabled() is false.
+ * They exist so we don't need #ifdefs around KVM-specific code that already
+ * checks kvm_enabled() properly.
+ */
+#define KVM_CPUID_FEATURES       0
+
+#endif  /* CONFIG_KVM */
+
 bool kvm_allows_irq0_override(void);
 bool kvm_has_smm(void);
 bool kvm_has_adjust_clock_stable(void);
-- 
1.8.3.1

[Qemu-devel] [PATCH v3 4/8] include/exec/poison: Mark CONFIG_KVM as poisoned, too

[Thomas Huth]

CONFIG_KVM is only defined for target-specific code, so nobody should
use it by accident in common code. To avoid such subtle bugs,
CONFIG_KVM is now marked as poisoned in common code. The header
include/sysemu/kvm.h is somewhat special since it is included
all over the place from common code, too, so we need some extra
logic via "#ifdef NEED_CPU_H" here to make sure that we can
compile all files without problems.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 hw/acpi/ich9.c        |  1 -
 include/exec/poison.h |  1 +
 include/sysemu/kvm.h  | 18 +++++++++++++-----
 3 files changed, 14 insertions(+), 6 deletions(-)

diff --git a/hw/acpi/ich9.c b/hw/acpi/ich9.c
index 5c279bb..c5d8646 100644
--- a/hw/acpi/ich9.c
+++ b/hw/acpi/ich9.c
@@ -33,7 +33,6 @@
 #include "sysemu/sysemu.h"
 #include "hw/acpi/acpi.h"
 #include "hw/acpi/tco.h"
-#include "sysemu/kvm.h"
 #include "exec/address-spaces.h"
 
 #include "hw/i386/ich9.h"
diff --git a/include/exec/poison.h b/include/exec/poison.h
index 7a025b2..32707cd 100644
--- a/include/exec/poison.h
+++ b/include/exec/poison.h
@@ -86,6 +86,7 @@
 
 #pragma GCC poison CONFIG_LINUX_USER
 #pragma GCC poison CONFIG_VHOST_NET
+#pragma GCC poison CONFIG_KVM
 
 #endif
 #endif
diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h
index ca40b6e..052e11f 100644
--- a/include/sysemu/kvm.h
+++ b/include/sysemu/kvm.h
@@ -19,11 +19,18 @@
 #include "exec/memattrs.h"
 #include "hw/irq.h"
 
-#ifdef CONFIG_KVM
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
+#ifdef NEED_CPU_H
+# ifdef CONFIG_KVM
+#  include <linux/kvm.h>
+#  include <linux/kvm_para.h>
+#  define CONFIG_KVM_IS_POSSIBLE
+# endif
+#else
+# define CONFIG_KVM_IS_POSSIBLE
 #endif
 
+#ifdef CONFIG_KVM_IS_POSSIBLE
+
 extern bool kvm_allowed;
 extern bool kvm_kernel_irqchip;
 extern bool kvm_split_irqchip;
@@ -40,7 +47,6 @@ extern bool kvm_direct_msi_allowed;
 extern bool kvm_ioeventfd_any_length_allowed;
 extern bool kvm_msi_use_devid;
 
-#if defined CONFIG_KVM || !defined NEED_CPU_H
 #define kvm_enabled()           (kvm_allowed)
 /**
  * kvm_irqchip_in_kernel:
@@ -163,6 +169,7 @@ extern bool kvm_msi_use_devid;
 #define kvm_msi_devid_required() (kvm_msi_use_devid)
 
 #else
+
 #define kvm_enabled()           (0)
 #define kvm_irqchip_in_kernel() (false)
 #define kvm_irqchip_is_split() (false)
@@ -178,7 +185,8 @@ extern bool kvm_msi_use_devid;
 #define kvm_direct_msi_enabled() (false)
 #define kvm_ioeventfd_any_length_enabled() (false)
 #define kvm_msi_devid_required() (false)
-#endif
+
+#endif  /* CONFIG_KVM_IS_POSSIBLE */
 
 struct kvm_run;
 struct kvm_lapic_state;
-- 
1.8.3.1

[Qemu-devel] [PATCH v3 5/8] cpu: Introduce a wrapper for tlb_flush() that can be used in common code

[Thomas Huth]

Commit 1f5c00cfdb8114c ("qom/cpu: move tlb_flush to cpu_common_reset")
moved the call to tlb_flush() from the target-specific reset handlers
into the common code qom/cpu.c file, and protected the call with
"#ifdef CONFIG_SOFTMMU" to avoid that it is called for linux-user
only targets. But since qom/cpu.c is common code, CONFIG_SOFTMMU is
*never* defined here, so the tlb_flush() was simply never executed
anymore. Fix it by introducing a wrapper for tlb_flush() in a file
that is re-compiled for each target, i.e. in translate-all.c.

Fixes: 1f5c00cfdb8114c1e3a13426588ceb64f82c9ddb
Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 include/exec/cpu-common.h | 2 ++
 qom/cpu.c                 | 5 ++---
 translate-all.c           | 8 ++++++++
 3 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h
index 4d45a72..74341b1 100644
--- a/include/exec/cpu-common.h
+++ b/include/exec/cpu-common.h
@@ -28,6 +28,8 @@ void qemu_init_cpu_list(void);
 void cpu_list_lock(void);
 void cpu_list_unlock(void);
 
+void tcg_flush_softmmu_tlb(CPUState *cs);
+
 #if !defined(CONFIG_USER_ONLY)
 
 enum device_endian {
diff --git a/qom/cpu.c b/qom/cpu.c
index 5069876..303eb42 100644
--- a/qom/cpu.c
+++ b/qom/cpu.c
@@ -26,6 +26,7 @@
 #include "qemu/notify.h"
 #include "qemu/log.h"
 #include "exec/log.h"
+#include "exec/cpu-common.h"
 #include "qemu/error-report.h"
 #include "sysemu/sysemu.h"
 #include "hw/qdev-properties.h"
@@ -296,9 +297,7 @@ static void cpu_common_reset(CPUState *cpu)
             atomic_set(&cpu->tb_jmp_cache[i], NULL);
         }
 
-#ifdef CONFIG_SOFTMMU
-        tlb_flush(cpu, 0);
-#endif
+        tcg_flush_softmmu_tlb(cpu);
     }
 }
 
diff --git a/translate-all.c b/translate-all.c
index b3ee876..a45480f 100644
--- a/translate-all.c
+++ b/translate-all.c
@@ -2219,3 +2219,11 @@ int page_unprotect(target_ulong address, uintptr_t pc)
     return 0;
 }
 #endif /* CONFIG_USER_ONLY */
+
+/* This is a wrapper for common code that can not use CONFIG_SOFTMMU */
+void tcg_flush_softmmu_tlb(CPUState *cs)
+{
+#ifdef CONFIG_SOFTMMU
+    tlb_flush(cs);
+#endif
+}
-- 
1.8.3.1

[Qemu-devel] [PATCH v3 6/8] include/exec/poison: Mark CONFIG_SOFTMMU as poisoned

[Thomas Huth]

CONFIG_SOFTMMU should never be used in common code, so mark
it as poisoned, too.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 include/exec/poison.h | 1 +
 include/qom/cpu.h     | 8 ++++++++
 2 files changed, 9 insertions(+)

diff --git a/include/exec/poison.h b/include/exec/poison.h
index 32707cd..41cd2eb 100644
--- a/include/exec/poison.h
+++ b/include/exec/poison.h
@@ -87,6 +87,7 @@
 #pragma GCC poison CONFIG_LINUX_USER
 #pragma GCC poison CONFIG_VHOST_NET
 #pragma GCC poison CONFIG_KVM
+#pragma GCC poison CONFIG_SOFTMMU
 
 #endif
 #endif
diff --git a/include/qom/cpu.h b/include/qom/cpu.h
index 89ddb68..c41e1e3 100644
--- a/include/qom/cpu.h
+++ b/include/qom/cpu.h
@@ -783,6 +783,8 @@ bool cpu_throttle_active(void);
  */
 int cpu_throttle_get_percentage(void);
 
+#ifdef NEED_CPU_H
+
 #ifndef CONFIG_USER_ONLY
 
 typedef void (*CPUInterruptHandler)(CPUState *, int);
@@ -829,6 +831,8 @@ static inline void cpu_unaligned_access(CPUState *cpu, vaddr addr,
 }
 #endif
 
+#endif /* NEED_CPU_H */
+
 /**
  * cpu_set_pc:
  * @cpu: The CPU to set the program counter for.
@@ -1005,6 +1009,8 @@ void cpu_exec_initfn(CPUState *cpu);
 void cpu_exec_realizefn(CPUState *cpu, Error **errp);
 void cpu_exec_unrealizefn(CPUState *cpu);
 
+#ifdef NEED_CPU_H
+
 #ifdef CONFIG_SOFTMMU
 extern const struct VMStateDescription vmstate_cpu_common;
 #else
@@ -1019,6 +1025,8 @@ extern const struct VMStateDescription vmstate_cpu_common;
     .offset = 0,                                                            \
 }
 
+#endif /* NEED_CPU_H */
+
 #define UNASSIGNED_CPU_INDEX -1
 
 #endif
-- 
1.8.3.1

[Qemu-devel] [PATCH v3 7/8] Makefile: Move bootdevice.o to common-obj-y

[Thomas Huth]

There does not seem to be any target specific code in this file, so
we can put it into "common-obj" instead of "obj" to compile it only
once for all targets.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 Makefile.objs   | 2 +-
 Makefile.target | 2 +-
 bootdevice.c    | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/Makefile.objs b/Makefile.objs
index 0575802..e5bdae9 100644
--- a/Makefile.objs
+++ b/Makefile.objs
@@ -40,7 +40,7 @@ io-obj-y = io/
 
 ifeq ($(CONFIG_SOFTMMU),y)
 common-obj-y = blockdev.o blockdev-nbd.o block/
-common-obj-y += iothread.o
+common-obj-y += bootdevice.o iothread.o
 common-obj-y += net/
 common-obj-y += qdev-monitor.o device-hotplug.o
 common-obj-$(CONFIG_WIN32) += os-win32.o
diff --git a/Makefile.target b/Makefile.target
index ce8dfe4..0759176 100644
--- a/Makefile.target
+++ b/Makefile.target
@@ -140,7 +140,7 @@ endif #CONFIG_BSD_USER
 # System emulator target
 ifdef CONFIG_SOFTMMU
 obj-y += arch_init.o cpus.o monitor.o gdbstub.o balloon.o ioport.o numa.o
-obj-y += qtest.o bootdevice.o
+obj-y += qtest.o
 obj-y += hw/
 obj-$(CONFIG_KVM) += kvm-all.o
 obj-y += memory.o cputlb.o
diff --git a/bootdevice.c b/bootdevice.c
index 33e3029..1141009 100644
--- a/bootdevice.c
+++ b/bootdevice.c
@@ -27,7 +27,7 @@
 #include "sysemu/sysemu.h"
 #include "qapi/visitor.h"
 #include "qemu/error-report.h"
-#include "hw/hw.h"
+#include "sysemu/reset.h"
 #include "hw/qdev-core.h"
 
 typedef struct FWBootEntry FWBootEntry;
-- 
1.8.3.1

[Qemu-devel] [PATCH v3 8/8] hw/misc/edu: Compile the edu device as common object

[Thomas Huth]

edu.c does not contain any target-specific code, so we can put
it into common-obj-y to compile it only once for all targets.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 hw/misc/Makefile.objs | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/hw/misc/Makefile.objs b/hw/misc/Makefile.objs
index 2019846..7fc4e41 100644
--- a/hw/misc/Makefile.objs
+++ b/hw/misc/Makefile.objs
@@ -6,6 +6,7 @@ common-obj-$(CONFIG_ISA_DEBUG) += debugexit.o
 common-obj-$(CONFIG_SGA) += sga.o
 common-obj-$(CONFIG_ISA_TESTDEV) += pc-testdev.o
 common-obj-$(CONFIG_PCI_TESTDEV) += pci-testdev.o
+common-obj-$(CONFIG_EDU) += edu.o
 
 common-obj-y += unimp.o
 
@@ -53,7 +54,6 @@ obj-$(CONFIG_MIPS_CPS) += mips_cpc.o
 obj-$(CONFIG_MIPS_ITU) += mips_itu.o
 
 obj-$(CONFIG_PVPANIC) += pvpanic.o
-obj-$(CONFIG_EDU) += edu.o
 obj-$(CONFIG_HYPERV_TESTDEV) += hyperv_testdev.o
 obj-$(CONFIG_AUX) += auxbus.o
 obj-$(CONFIG_ASPEED_SOC) += aspeed_scu.o aspeed_sdmc.o
-- 
1.8.3.1

Re: [Qemu-devel] [PATCH v3 1/8] include/exec/poison: Add missing TARGET defines

[Richard Henderson]

On 06/19/2017 10:52 PM, Thomas Huth wrote:
> Since we've got some new CPU targets in QEMU during the last months
> and years, we've got some new TARGET_xxx defines now which should
> be marked as poisoned for common code.
> 
> Signed-off-by: Thomas Huth<thuth@redhat.com>
> ---
>   include/exec/poison.h | 16 ++++++++++++++++
>   1 file changed, 16 insertions(+)

Reviewed-by: Richard Henderson <rth@twiddle.net>


r~

Re: [Qemu-devel] [PATCH v3 2/8] include/exec/poison: Mark some CONFIG defines as poisoned, too

[Richard Henderson]

On 06/19/2017 10:52 PM, Thomas Huth wrote:
> These are defined in config-target.h and thus should never be
> used in common code.
> 
> Signed-off-by: Thomas Huth<thuth@redhat.com>
> ---
>   include/exec/poison.h | 21 +++++++++++++++++++++
>   1 file changed, 21 insertions(+)

Reviewed-by: Richard Henderson <rth@twiddle.net>


r~

Re: [Qemu-devel] [PATCH v3 5/8] cpu: Introduce a wrapper for tlb_flush() that can be used in common code

[Richard Henderson]

On 06/19/2017 10:52 PM, Thomas Huth wrote:
> Commit 1f5c00cfdb8114c ("qom/cpu: move tlb_flush to cpu_common_reset")
> moved the call to tlb_flush() from the target-specific reset handlers
> into the common code qom/cpu.c file, and protected the call with
> "#ifdef CONFIG_SOFTMMU" to avoid that it is called for linux-user
> only targets. But since qom/cpu.c is common code, CONFIG_SOFTMMU is
> *never*  defined here, so the tlb_flush() was simply never executed
> anymore. Fix it by introducing a wrapper for tlb_flush() in a file
> that is re-compiled for each target, i.e. in translate-all.c.
> 
> Fixes: 1f5c00cfdb8114c1e3a13426588ceb64f82c9ddb
> Signed-off-by: Thomas Huth<thuth@redhat.com>
> ---
>   include/exec/cpu-common.h | 2 ++
>   qom/cpu.c                 | 5 ++---
>   translate-all.c           | 8 ++++++++
>   3 files changed, 12 insertions(+), 3 deletions(-)

Reviewed-by: Richard Henderson <rth@twiddle.net>


r~

Re: [Qemu-devel] [PATCH v3 6/8] include/exec/poison: Mark CONFIG_SOFTMMU as poisoned

[Richard Henderson]

On 06/19/2017 10:52 PM, Thomas Huth wrote:
> CONFIG_SOFTMMU should never be used in common code, so mark
> it as poisoned, too.
> 
> Signed-off-by: Thomas Huth<thuth@redhat.com>
> ---
>   include/exec/poison.h | 1 +
>   include/qom/cpu.h     | 8 ++++++++
>   2 files changed, 9 insertions(+)

Reviewed-by: Richard Henderson <rth@twiddle.net>


r~

Re: [Qemu-devel] [PATCH v3 8/8] hw/misc/edu: Compile the edu device as common object

[Richard Henderson]

On 06/19/2017 10:52 PM, Thomas Huth wrote:
> edu.c does not contain any target-specific code, so we can put
> it into common-obj-y to compile it only once for all targets.
> 
> Signed-off-by: Thomas Huth <thuth@redhat.com>

Isn't CONFIG_EDU defined by the target, via pci.conf?
Would it ever be set generically?


r~

Re: [Qemu-devel] [PATCH v3 8/8] hw/misc/edu: Compile the edu device as common object

[Thomas Huth]

On 20.06.2017 20:14, Richard Henderson wrote:
> On 06/19/2017 10:52 PM, Thomas Huth wrote:
>> edu.c does not contain any target-specific code, so we can put
>> it into common-obj-y to compile it only once for all targets.
>>
>> Signed-off-by: Thomas Huth <thuth@redhat.com>
> 
> Isn't CONFIG_EDU defined by the target, via pci.conf?
> Would it ever be set generically?

This is about Makefiles, not #defines in C sources ... so if I
understand our build-system correctly, the variable is set for the
Makefiles of the targets that use pci.mak. As far as I see, it gets
compiled correctly here.
And we've got plenty of those common-obj-$(CONFIG_xxx) statements
elsewhere in this file (e.g. for CONFIG_PCI_TESTDEV which is also
declared in default-configs/pci.mak), so this is just normal, no problem
here, as far as I can tell.

 Thomas

Re: [Qemu-devel] [PATCH v3 8/8] hw/misc/edu: Compile the edu device as common object

[Paolo Bonzini]

----- Original Message -----
> From: "Thomas Huth" <thuth@redhat.com>
> To: "Richard Henderson" <rth@twiddle.net>, qemu-devel@nongnu.org, "Paolo Bonzini" <pbonzini@redhat.com>
> Sent: Wednesday, June 21, 2017 3:40:31 AM
> Subject: Re: [PATCH v3 8/8] hw/misc/edu: Compile the edu device as common object
> 
> On 20.06.2017 20:14, Richard Henderson wrote:
> > On 06/19/2017 10:52 PM, Thomas Huth wrote:
> >> edu.c does not contain any target-specific code, so we can put
> >> it into common-obj-y to compile it only once for all targets.
> >>
> >> Signed-off-by: Thomas Huth <thuth@redhat.com>
> > 
> > Isn't CONFIG_EDU defined by the target, via pci.conf?
> > Would it ever be set generically?
> 
> This is about Makefiles, not #defines in C sources ... so if I
> understand our build-system correctly, the variable is set for the
> Makefiles of the targets that use pci.mak. As far as I see, it gets
> compiled correctly here.
> And we've got plenty of those common-obj-$(CONFIG_xxx) statements
> elsewhere in this file (e.g. for CONFIG_PCI_TESTDEV which is also
> declared in default-configs/pci.mak), so this is just normal, no problem
> here, as far as I can tell.

Yes, the variable is evaluated once during the main directory build,
where it comes from config-all-devices.mak.  Then it is evaluated again
during the per-target build, where it comes from target-specific
configuration.

config-all-devices.mak should be documented in docs/devel/build-system.txt.

Paolo
Paolo