[PATCH 0/3] linux-user, configure: fix CPU canonicalization

[PATCH 0/3] linux-user, configure: fix CPU canonicalization

[Paolo Bonzini]

The CPU model has to be canonicalized to what Meson wants in the cross
file, to what Linux uses for its asm-$ARCH directories, and to what
QEMU uses for its user-mode emulation host/$ARCH directories.  Do
all three in a single case statement, and check that the Linux and
QEMU directories actually exist.

At a small cost in repeated lines, this ensures that there are no hidden
ordering requirements between the case statements.  In particular, commit
89e5b7935e9 ("configure: Fix linux-user host detection for riscv64",
2023-08-06) broke ppc64le because it assigned host_arch based on a
non-canonicalized version of $cpu.

While doing this, I noticed that linux-user won't work on x32, alpha
and 32-bit s390 these days, due to missing common-user/ fragments.
The first two patches clean up the directories.

Paolo Bonzini (3):
  linux-user: fix detection for x32
  linux-user: cleanup unused linux-user/include/host directories
  configure: unify case statements for CPU canonicalization

 configure                                   | 157 +++++++++++---------
 linux-user/include/host/alpha/host-signal.h |  55 -------
 linux-user/include/host/s390/host-signal.h  | 138 -----------------
 linux-user/include/host/x32/host-signal.h   |   1 -
 4 files changed, 87 insertions(+), 264 deletions(-)
 delete mode 100644 linux-user/include/host/alpha/host-signal.h
 delete mode 100644 linux-user/include/host/s390/host-signal.h
 delete mode 100644 linux-user/include/host/x32/host-signal.h

-- 
2.41.0

[PATCH 1/3] configure: fix detection for x32 linux-user

[Paolo Bonzini]

x32 uses the same signal handling fragments as x86_64, since host_arch
is set to x86_64 when Meson runs.  Remove the unnecessary forwarder and
set the host_arch variable properly in configure.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 configure                                 | 2 ++
 linux-user/include/host/x32/host-signal.h | 1 -
 2 files changed, 2 insertions(+), 1 deletion(-)
 delete mode 100644 linux-user/include/host/x32/host-signal.h

diff --git a/configure b/configure
index 98dc78280e6..484d38d81f4 100755
--- a/configure
+++ b/configure
@@ -472,6 +472,8 @@ fi
 case "$cpu" in
   riscv*)
     host_arch=riscv ;;
+  x32)
+    host_arch=x86_64 ;;
   *)
     host_arch="$cpu" ;;
 esac
diff --git a/linux-user/include/host/x32/host-signal.h b/linux-user/include/host/x32/host-signal.h
deleted file mode 100644
index 26800591d3b..00000000000
--- a/linux-user/include/host/x32/host-signal.h
+++ /dev/null
@@ -1 +0,0 @@
-#include "../x86_64/host-signal.h"
-- 
2.41.0

Re: [PATCH 1/3] configure: fix detection for x32 linux-user

[Philippe Mathieu-Daudé]

On 7/8/23 11:48, Paolo Bonzini wrote:
> x32 uses the same signal handling fragments as x86_64, since host_arch
> is set to x86_64 when Meson runs.  Remove the unnecessary forwarder and
> set the host_arch variable properly in configure.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>   configure                                 | 2 ++
>   linux-user/include/host/x32/host-signal.h | 1 -
>   2 files changed, 2 insertions(+), 1 deletion(-)
>   delete mode 100644 linux-user/include/host/x32/host-signal.h

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

[PATCH 2/3] linux-user: cleanup unused linux-user/include/host directories

[Paolo Bonzini]

Alpha and 31-bit s390 lack the assembly fragment to handle signals
occurring at the same time as system calls, so they cannot run
linux-user emulation anymore.  Drop the host-signal.h files for
them.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 linux-user/include/host/alpha/host-signal.h |  55 --------
 linux-user/include/host/s390/host-signal.h  | 138 --------------------
 2 files changed, 193 deletions(-)
 delete mode 100644 linux-user/include/host/alpha/host-signal.h
 delete mode 100644 linux-user/include/host/s390/host-signal.h

diff --git a/linux-user/include/host/alpha/host-signal.h b/linux-user/include/host/alpha/host-signal.h
deleted file mode 100644
index 4f9e2abc4b0..00000000000
--- a/linux-user/include/host/alpha/host-signal.h
+++ /dev/null
@@ -1,55 +0,0 @@
-/*
- * host-signal.h: signal info dependent on the host architecture
- *
- * Copyright (c) 2003-2005 Fabrice Bellard
- * Copyright (c) 2021 Linaro Limited
- *
- * This work is licensed under the terms of the GNU LGPL, version 2.1 or later.
- * See the COPYING file in the top-level directory.
- */
-
-#ifndef ALPHA_HOST_SIGNAL_H
-#define ALPHA_HOST_SIGNAL_H
-
-/* The third argument to a SA_SIGINFO handler is ucontext_t. */
-typedef ucontext_t host_sigcontext;
-
-static inline uintptr_t host_signal_pc(host_sigcontext *uc)
-{
-    return uc->uc_mcontext.sc_pc;
-}
-
-static inline void host_signal_set_pc(host_sigcontext *uc, uintptr_t pc)
-{
-    uc->uc_mcontext.sc_pc = pc;
-}
-
-static inline void *host_signal_mask(host_sigcontext *uc)
-{
-    return &uc->uc_sigmask;
-}
-
-static inline bool host_signal_write(siginfo_t *info, host_sigcontext *uc)
-{
-    uint32_t *pc = (uint32_t *)host_signal_pc(uc);
-    uint32_t insn = *pc;
-
-    /* XXX: need kernel patch to get write flag faster */
-    switch (insn >> 26) {
-    case 0x0d: /* stw */
-    case 0x0e: /* stb */
-    case 0x0f: /* stq_u */
-    case 0x24: /* stf */
-    case 0x25: /* stg */
-    case 0x26: /* sts */
-    case 0x27: /* stt */
-    case 0x2c: /* stl */
-    case 0x2d: /* stq */
-    case 0x2e: /* stl_c */
-    case 0x2f: /* stq_c */
-        return true;
-    }
-    return false;
-}
-
-#endif
diff --git a/linux-user/include/host/s390/host-signal.h b/linux-user/include/host/s390/host-signal.h
deleted file mode 100644
index e6d3ec26dc7..00000000000
--- a/linux-user/include/host/s390/host-signal.h
+++ /dev/null
@@ -1,138 +0,0 @@
-/*
- * host-signal.h: signal info dependent on the host architecture
- *
- * Copyright (c) 2003-2005 Fabrice Bellard
- * Copyright (c) 2021 Linaro Limited
- *
- * This work is licensed under the terms of the GNU LGPL, version 2.1 or later.
- * See the COPYING file in the top-level directory.
- */
-
-#ifndef S390_HOST_SIGNAL_H
-#define S390_HOST_SIGNAL_H
-
-/* The third argument to a SA_SIGINFO handler is ucontext_t. */
-typedef ucontext_t host_sigcontext;
-
-static inline uintptr_t host_signal_pc(host_sigcontext *uc)
-{
-    return uc->uc_mcontext.psw.addr;
-}
-
-static inline void host_signal_set_pc(host_sigcontext *uc, uintptr_t pc)
-{
-    uc->uc_mcontext.psw.addr = pc;
-}
-
-static inline void *host_signal_mask(host_sigcontext *uc)
-{
-    return &uc->uc_sigmask;
-}
-
-static inline bool host_signal_write(siginfo_t *info, host_sigcontext *uc)
-{
-    uint16_t *pinsn = (uint16_t *)host_signal_pc(uc);
-
-    /*
-     * ??? On linux, the non-rt signal handler has 4 (!) arguments instead
-     * of the normal 2 arguments.  The 4th argument contains the "Translation-
-     * Exception Identification for DAT Exceptions" from the hardware (aka
-     * "int_parm_long"), which does in fact contain the is_write value.
-     * The rt signal handler, as far as I can tell, does not give this value
-     * at all.  Not that we could get to it from here even if it were.
-     * So fall back to parsing instructions.  Treat read-modify-write ones as
-     * writes, which is not fully correct, but for tracking self-modifying code
-     * this is better than treating them as reads.  Checking si_addr page flags
-     * might be a viable improvement, albeit a racy one.
-     */
-    /* ??? This is not even close to complete.  */
-    switch (pinsn[0] >> 8) {
-    case 0x50: /* ST */
-    case 0x42: /* STC */
-    case 0x40: /* STH */
-    case 0x44: /* EX */
-    case 0xba: /* CS */
-    case 0xbb: /* CDS */
-        return true;
-    case 0xc4: /* RIL format insns */
-        switch (pinsn[0] & 0xf) {
-        case 0xf: /* STRL */
-        case 0xb: /* STGRL */
-        case 0x7: /* STHRL */
-            return true;
-        }
-        break;
-    case 0xc6: /* RIL-b format insns */
-        switch (pinsn[0] & 0xf) {
-        case 0x0: /* EXRL */
-            return true;
-        }
-        break;
-    case 0xc8: /* SSF format insns */
-        switch (pinsn[0] & 0xf) {
-        case 0x2: /* CSST */
-            return true;
-        }
-        break;
-    case 0xe3: /* RXY format insns */
-        switch (pinsn[2] & 0xff) {
-        case 0x50: /* STY */
-        case 0x24: /* STG */
-        case 0x72: /* STCY */
-        case 0x70: /* STHY */
-        case 0x8e: /* STPQ */
-        case 0x3f: /* STRVH */
-        case 0x3e: /* STRV */
-        case 0x2f: /* STRVG */
-            return true;
-        }
-        break;
-    case 0xe6:
-        switch (pinsn[2] & 0xff) {
-        case 0x09: /* VSTEBRH */
-        case 0x0a: /* VSTEBRG */
-        case 0x0b: /* VSTEBRF */
-        case 0x0e: /* VSTBR */
-        case 0x0f: /* VSTER */
-        case 0x3f: /* VSTRLR */
-            return true;
-        }
-        break;
-    case 0xe7:
-        switch (pinsn[2] & 0xff) {
-        case 0x08: /* VSTEB */
-        case 0x09: /* VSTEH */
-        case 0x0a: /* VSTEG */
-        case 0x0b: /* VSTEF */
-        case 0x0e: /* VST */
-        case 0x1a: /* VSCEG */
-        case 0x1b: /* VSCEF */
-        case 0x3e: /* VSTM */
-        case 0x3f: /* VSTL */
-            return true;
-        }
-        break;
-    case 0xeb: /* RSY format insns */
-        switch (pinsn[2] & 0xff) {
-        case 0x14: /* CSY */
-        case 0x30: /* CSG */
-        case 0x31: /* CDSY */
-        case 0x3e: /* CDSG */
-        case 0xe4: /* LANG */
-        case 0xe6: /* LAOG */
-        case 0xe7: /* LAXG */
-        case 0xe8: /* LAAG */
-        case 0xea: /* LAALG */
-        case 0xf4: /* LAN */
-        case 0xf6: /* LAO */
-        case 0xf7: /* LAX */
-        case 0xfa: /* LAAL */
-        case 0xf8: /* LAA */
-            return true;
-        }
-        break;
-    }
-    return false;
-}
-
-#endif
-- 
2.41.0

Re: [PATCH 2/3] linux-user: cleanup unused linux-user/include/host directories

[Ilya Leoshkevich]

On Mon, 2023-08-07 at 11:48 +0200, Paolo Bonzini wrote:
> Alpha and 31-bit s390 lack the assembly fragment to handle signals
> occurring at the same time as system calls, so they cannot run
> linux-user emulation anymore.  Drop the host-signal.h files for
> them.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> ---
>  linux-user/include/host/alpha/host-signal.h |  55 --------
>  linux-user/include/host/s390/host-signal.h  | 138 ------------------
> --
>  2 files changed, 193 deletions(-)
>  delete mode 100644 linux-user/include/host/alpha/host-signal.h
>  delete mode 100644 linux-user/include/host/s390/host-signal.h

s390/host-signal.h is included by s390x/host-signal.h, so on a (64-bit)
s390x host I get:

In file included from ../linux-user/signal.c:33:
../linux-user/include/host/s390x/host-signal.h:1:10: fatal error:
../s390/host-signal.h: No such file or directory
    1 | #include "../s390/host-signal.h"

Re: [PATCH 2/3] linux-user: cleanup unused linux-user/include/host directories

[Paolo Bonzini]

[-- Attachment #1: Type: text/plain, Size: 1176 bytes --]

Yes, found the same in CI. I will send V2 tomorrow.

Paolo

Il lun 7 ago 2023, 19:18 Ilya Leoshkevich <iii@linux.ibm.com> ha scritto:

> On Mon, 2023-08-07 at 11:48 +0200, Paolo Bonzini wrote:
> > Alpha and 31-bit s390 lack the assembly fragment to handle signals
> > occurring at the same time as system calls, so they cannot run
> > linux-user emulation anymore.  Drop the host-signal.h files for
> > them.
> >
> > Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> > Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> > ---
> >  linux-user/include/host/alpha/host-signal.h |  55 --------
> >  linux-user/include/host/s390/host-signal.h  | 138 ------------------
> > --
> >  2 files changed, 193 deletions(-)
> >  delete mode 100644 linux-user/include/host/alpha/host-signal.h
> >  delete mode 100644 linux-user/include/host/s390/host-signal.h
>
> s390/host-signal.h is included by s390x/host-signal.h, so on a (64-bit)
> s390x host I get:
>
> In file included from ../linux-user/signal.c:33:
> ../linux-user/include/host/s390x/host-signal.h:1:10: fatal error:
> ../s390/host-signal.h: No such file or directory
>     1 | #include "../s390/host-signal.h"
>
>

[-- Attachment #2: Type: text/html, Size: 1786 bytes --]

[PATCH 3/3] configure: unify case statements for CPU canonicalization

[Paolo Bonzini]

The CPU model has to be canonicalized to what Meson wants in the cross
file, to what Linux uses for its asm-$ARCH directories, and to what
QEMU uses for its user-mode emulation host/$ARCH directories.  Do
all three in a single case statement, and check that the Linux and
QEMU directories actually exist.

At a small cost in repeated lines, this ensures that there are no hidden
ordering requirements between the case statements.  In particular, commit
89e5b7935e9 ("configure: Fix linux-user host detection for riscv64",
2023-08-06) broke ppc64le because it assigned host_arch based on a
non-canonicalized version of $cpu.

Reported-by: Joel Stanley <joel@jms.id.au>
Fixes: 89e5b7935e9 ("configure: Fix linux-user host detection for riscv64", 2023-08-06)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 configure | 159 +++++++++++++++++++++++++++++-------------------------
 1 file changed, 87 insertions(+), 72 deletions(-)

diff --git a/configure b/configure
index 484d38d81f4..fb213412d7c 100755
--- a/configure
+++ b/configure
@@ -469,59 +469,104 @@ else
   echo "WARNING: unrecognized host CPU, proceeding with 'uname -m' output '$cpu'"
 fi
 
-case "$cpu" in
-  riscv*)
-    host_arch=riscv ;;
-  x32)
-    host_arch=x86_64 ;;
-  *)
-    host_arch="$cpu" ;;
-esac
-
-# Normalise host CPU name and set multilib cflags.  The canonicalization
-# isn't really necessary, because the architectures that we check for
-# should not hit the 'uname -m' case, but better safe than sorry.
+# Normalise host CPU name to the values used by Meson cross files and in source
+# directories, and set multilib cflags.  The canonicalization isn't really
+# necessary, because the architectures that we check for should not hit the
+# 'uname -m' case, but better safe than sorry.
+#
 # Note that this case should only have supported host CPUs, not guests.
+# Please keep it sorted.
+host_arch=
+linux_arch=
 case "$cpu" in
+  aarch64)
+    host_arch=aarch64
+    linux_arch=arm64 ;;
+
   armv*b|armv*l|arm)
     cpu="arm" ;;
+    host_arch=arm
+    linux_arch=arm ;;
 
   i386|i486|i586|i686)
     cpu="i386"
+    host_arch=i386
+    linux_arch=x86
     CPU_CFLAGS="-m32" ;;
+
+  loongarch*)
+    cpu=loongarch64
+    host_arch=loongarch64 ;;
+
+  mips64*)
+    cpu=mips64
+    host_arch=mips
+    linux_arch=mips ;;
+  mips*)
+    cpu=mips
+    host_arch=mips
+    linux_arch=mips ;;
+
+  ppc)
+    host_arch=ppc
+    linux_arch=powerpc
+    CPU_CFLAGS="-m32" ;;
+  ppc64)
+    host_arch=ppc64
+    linux_arch=powerpc
+    CPU_CFLAGS="-m64 -mbig-endian" ;;
+  ppc64le)
+    cpu=ppc64
+    host_arch=ppc64
+    linux_arch=powerpc
+    CPU_CFLAGS="-m64 -mlittle-endian" ;;
+
+  riscv32 | riscv64)
+    host_arch=riscv
+    linux_arch=riscv
+    ;;
+
+  s390)
+    linux_arch=s390
+    CPU_CFLAGS="-m31" ;;
+  s390x)
+    host_arch=s390x
+    linux_arch=s390
+    CPU_CFLAGS="-m64" ;;
+
+  sparc|sun4[cdmuv])
+    cpu=sparc
+    CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc" ;;
+  sparc64)
+    host_arch=sparc64
+    CPU_CFLAGS="-m64 -mcpu=ultrasparc" ;;
+
   x32)
     cpu="x86_64"
+    host_arch=x86_64
+    linux_arch=x86
     CPU_CFLAGS="-mx32" ;;
   x86_64|amd64)
     cpu="x86_64"
+    host_arch=x86_64
+    linux_arch=x86
     # ??? Only extremely old AMD cpus do not have cmpxchg16b.
     # If we truly care, we should simply detect this case at
     # runtime and generate the fallback to serial emulation.
     CPU_CFLAGS="-m64 -mcx16" ;;
-
-  mips*)
-    cpu="mips" ;;
-
-  ppc)
-    CPU_CFLAGS="-m32" ;;
-  ppc64)
-    CPU_CFLAGS="-m64 -mbig-endian" ;;
-  ppc64le)
-    cpu="ppc64"
-    CPU_CFLAGS="-m64 -mlittle-endian" ;;
-
-  s390)
-    CPU_CFLAGS="-m31" ;;
-  s390x)
-    CPU_CFLAGS="-m64" ;;
-
-  sparc|sun4[cdmuv])
-    cpu="sparc"
-    CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc" ;;
-  sparc64)
-    CPU_CFLAGS="-m64 -mcpu=ultrasparc" ;;
 esac
 
+if test -n "$host_arch" && {
+    ! test -d "$source_path/linux-user/include/host/$host_arch" ||
+    ! test -d "$source_path/common-user/host/$host_arch"; }; then
+    error_exit "linux-user/include/host/$host_arch does not exist." \
+       "This is a bug in the configure script, please report it."
+fi
+if test -n "$linux_arch" && ! test -d "$source_path/linux-headers/asm-$linux_arch"; then
+    error_exit "linux-headers/asm-$linux_arch does not exist." \
+       "This is a bug in the configure script, please report it."
+fi
+
 check_py_version() {
     # We require python >= 3.7.
     # NB: a True python conditional creates a non-zero return code (Failure)
@@ -812,7 +857,7 @@ default_target_list=""
 mak_wilds=""
 
 if [ "$linux_user" != no ]; then
-    if [ "$targetos" = linux ] && [ -d "$source_path/linux-user/include/host/$host_arch" ]; then
+    if [ "$targetos" = linux ] && [ -n "$host_arch" ]; then
         linux_user=yes
     elif [ "$linux_user" = yes ]; then
         error_exit "linux-user not supported on this architecture"
@@ -1717,40 +1762,10 @@ echo "PKG_CONFIG=${pkg_config}" >> $config_host_mak
 echo "CC=$cc" >> $config_host_mak
 echo "EXESUF=$EXESUF" >> $config_host_mak
 
-# use included Linux headers
-if test "$linux" = "yes" ; then
+# use included Linux headers for KVM architectures
+if test "$linux" = "yes" && test -n "$linux_arch"; then
   mkdir -p linux-headers
-  case "$cpu" in
-  i386|x86_64)
-    linux_arch=x86
-    ;;
-  ppc|ppc64)
-    linux_arch=powerpc
-    ;;
-  s390x)
-    linux_arch=s390
-    ;;
-  aarch64)
-    linux_arch=arm64
-    ;;
-  loongarch*)
-    linux_arch=loongarch
-    ;;
-  mips64)
-    linux_arch=mips
-    ;;
-  riscv32|riscv64)
-    linux_arch=riscv
-    ;;
-  *)
-    # For most CPUs the kernel architecture name and QEMU CPU name match.
-    linux_arch="$cpu"
-    ;;
-  esac
-    # For non-KVM architectures we will not have asm headers
-    if [ -e "$source_path/linux-headers/asm-$linux_arch" ]; then
-      symlink "$source_path/linux-headers/asm-$linux_arch" linux-headers/asm
-    fi
+  symlink "$source_path/linux-headers/asm-$linux_arch" linux-headers/asm
 fi
 
 for target in $target_list; do
-- 
2.41.0

Re: [PATCH 3/3] configure: unify case statements for CPU canonicalization

[Paolo Bonzini]

On 8/7/23 11:48, Paolo Bonzini wrote:
>     armv*b|armv*l|arm)
>       cpu="arm" ;;

Oops, sent a stale patchset. The only difference is this extra ;;

Paolo

Re: [PATCH 3/3] configure: unify case statements for CPU canonicalization

[Peter Maydell]

On Mon, 7 Aug 2023 at 10:49, Paolo Bonzini <pbonzini@redhat.com> wrote:
>
> The CPU model has to be canonicalized to what Meson wants in the cross
> file, to what Linux uses for its asm-$ARCH directories, and to what
> QEMU uses for its user-mode emulation host/$ARCH directories.  Do
> all three in a single case statement, and check that the Linux and
> QEMU directories actually exist.
>
> At a small cost in repeated lines, this ensures that there are no hidden
> ordering requirements between the case statements.  In particular, commit
> 89e5b7935e9 ("configure: Fix linux-user host detection for riscv64",
> 2023-08-06) broke ppc64le because it assigned host_arch based on a
> non-canonicalized version of $cpu.
>
> Reported-by: Joel Stanley <joel@jms.id.au>
> Fixes: 89e5b7935e9 ("configure: Fix linux-user host detection for riscv64", 2023-08-06)
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  configure | 159 +++++++++++++++++++++++++++++-------------------------
>  1 file changed, 87 insertions(+), 72 deletions(-)
>
> diff --git a/configure b/configure
> index 484d38d81f4..fb213412d7c 100755
> --- a/configure
> +++ b/configure
> @@ -469,59 +469,104 @@ else
>    echo "WARNING: unrecognized host CPU, proceeding with 'uname -m' output '$cpu'"
>  fi
>
> -case "$cpu" in
> -  riscv*)
> -    host_arch=riscv ;;
> -  x32)
> -    host_arch=x86_64 ;;
> -  *)
> -    host_arch="$cpu" ;;
> -esac
> -
> -# Normalise host CPU name and set multilib cflags.  The canonicalization
> -# isn't really necessary, because the architectures that we check for
> -# should not hit the 'uname -m' case, but better safe than sorry.
> +# Normalise host CPU name to the values used by Meson cross files and in source
> +# directories, and set multilib cflags.  The canonicalization isn't really
> +# necessary, because the architectures that we check for should not hit the
> +# 'uname -m' case, but better safe than sorry.
> +#
>  # Note that this case should only have supported host CPUs, not guests.
> +# Please keep it sorted.
> +host_arch=
> +linux_arch=
>  case "$cpu" in
> +  aarch64)
> +    host_arch=aarch64
> +    linux_arch=arm64 ;;
> +
>    armv*b|armv*l|arm)
>      cpu="arm" ;;
> +    host_arch=arm
> +    linux_arch=arm ;;
>
>    i386|i486|i586|i686)
>      cpu="i386"
> +    host_arch=i386
> +    linux_arch=x86
>      CPU_CFLAGS="-m32" ;;
> +
> +  loongarch*)
> +    cpu=loongarch64
> +    host_arch=loongarch64 ;;
> +
> +  mips64*)
> +    cpu=mips64
> +    host_arch=mips
> +    linux_arch=mips ;;
> +  mips*)
> +    cpu=mips
> +    host_arch=mips
> +    linux_arch=mips ;;
> +
> +  ppc)
> +    host_arch=ppc
> +    linux_arch=powerpc
> +    CPU_CFLAGS="-m32" ;;
> +  ppc64)
> +    host_arch=ppc64
> +    linux_arch=powerpc
> +    CPU_CFLAGS="-m64 -mbig-endian" ;;
> +  ppc64le)
> +    cpu=ppc64
> +    host_arch=ppc64
> +    linux_arch=powerpc
> +    CPU_CFLAGS="-m64 -mlittle-endian" ;;
> +
> +  riscv32 | riscv64)
> +    host_arch=riscv
> +    linux_arch=riscv
> +    ;;

Can we be consistent within this case statement about whether
the ';;' is on its own line or at the end of the last line of
the case ? We are not fully consistent within the entire
configure script, but mostly we put it on a line of its
own. (I prefer that personally.)

thanks
-- PMM

Re: [PATCH 3/3] configure: unify case statements for CPU canonicalization

[Paolo Bonzini]

[-- Attachment #1: Type: text/plain, Size: 432 bytes --]

Il lun 7 ago 2023, 15:45 Peter Maydell <peter.maydell@linaro.org> ha
scritto:

> Can we be consistent within this case statement about whether
> the ';;' is on its own line or at the end of the last line of
> the case ? We are not fully consistent within the entire
> configure script, but mostly we put it on a line of its
> own. (I prefer that personally.)
>

Yes, especially given the nicer diffs.

Paolo


> thanks
> -- PMM
>
>

[-- Attachment #2: Type: text/html, Size: 1034 bytes --]

Re: [PATCH 0/3] linux-user, configure: fix CPU canonicalization

[Richard Henderson]

On 8/7/23 02:48, Paolo Bonzini wrote:
> Paolo Bonzini (3):
>    linux-user: fix detection for x32
>    linux-user: cleanup unused linux-user/include/host directories
>    configure: unify case statements for CPU canonicalization

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

Thanks for cleaning this up.


r~