All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Radim Krčmář" <rkrcmar@redhat.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: kvm@vger.kernel.org, Drew Jones <drjones@redhat.com>,
	Laurent Vivier <lvivier@redhat.com>,
	Thomas Huth <thuth@redhat.com>,
	David Hildenbrand <david@redhat.com>
Subject: Re: [kvm-unit-tests PATCH 5/5] arch/run: unify accelerator detection
Date: Thu, 29 Jun 2017 15:46:57 +0200	[thread overview]
Message-ID: <20170629134657.GA14932@potion> (raw)
In-Reply-To: <20170629124549.GB10661@potion>

2017-06-29 14:45+0200, Radim Krčmář:
> 2017-06-28 22:48+0200, Paolo Bonzini:
> > On 28/06/2017 22:08, Radim Krčmář wrote:
> > > -if [ "$ACCEL" = "kvm" ] && [ "$kvm_available" != "yes" ]; then
> > > -	echo "KVM is needed, but not available on this host"
> > > +ACCEL=$(get_accel '([ "$HOST" = "arm" ] && [ "$ARCH" = "arm" ]) || [ "$HOST" = "aarch64" ]') ||
> > >  	exit 2
> > 
> > Should x86 do the same as ARM (follow-up patch)?
> 
> Yeah, I'll prepare it when we've agreed on the format.
> 
> > And maybe the test could move from an argument of get_accel to a
> > function "arch_kvm_available" that returns 0 if available and 1 if not?
> 
> Right, the arm condition needs to be fixed as
> 
>   ([ "$HOST" = "arm" ] && [ "$ARCH" = "arm" ]) ||
>   ([ "$HOST" = "aarch64" ] && [ "$ARCH" = arm -o "$ARCH" = arm64 ])
> 
> and it would look even worse in the argument ...
> 
> Would it be ok like this?

No, we can make a better use of the $HOST variable:

diff --git a/arm/run b/arm/run
index f2e71eca022d..1245d42edae7 100755
--- a/arm/run
+++ b/arm/run
@@ -10,7 +10,7 @@ if [ -z "$STANDALONE" ]; then
 fi
 processor="$PROCESSOR"
 
-ACCEL=$(get_accel '([ "$HOST" = "arm" ] && [ "$ARCH" = "arm" ]) || [ "$HOST" = "aarch64" ]') ||
+ACCEL=$(get_accel) ||
 	exit 2
 
 qemu=$(search_qemu_binary) ||
diff --git a/powerpc/run b/powerpc/run
index 55b987c5efc5..58921e81eb8f 100755
--- a/powerpc/run
+++ b/powerpc/run
@@ -9,7 +9,7 @@ if [ -z "$STANDALONE" ]; then
 	source scripts/arch-run.bash
 fi
 
-ACCEL=$(get_accel '[ "$HOST" = "ppc64" ] && [ "$ARCH" = "ppc64" ]') ||
+ACCEL=$(get_accel) ||
 	exit 2
 
 qemu=$(search_qemu_binary) ||
diff --git a/s390x/run b/s390x/run
index 67d6e88ea7e4..2242ebb79fbc 100755
--- a/s390x/run
+++ b/s390x/run
@@ -9,7 +9,7 @@ if [ -z "$STANDALONE" ]; then
 	source scripts/arch-run.bash
 fi
 
-ACCEL=$(get_accel '[ "$HOST" = "s390x" ] && [ "$ARCH" = "s390x" ]') ||
+ACCEL=$(get_accel) ||
 	exit 2
 
 qemu=$(search_qemu_binary) ||
diff --git a/scripts/arch-run.bash b/scripts/arch-run.bash
index a696652c61b1..77071216ff99 100644
--- a/scripts/arch-run.bash
+++ b/scripts/arch-run.bash
@@ -273,22 +273,26 @@ trap_exit_push ()
 	trap -- "$1; $old_exit" EXIT
 }
 
+kvm_available ()
+{
+	[ -c /dev/kvm ] ||
+		return 1
+
+	[ "$HOST" = "$ARCH_NAME" ] ||
+		[ "$HOST" = aarch64 -a "$ARCH" = arm ] ||
+		[ "$HOST" = x86_64 -a "$ARCH" = i386 ]
+}
+
 get_accel ()
 {
-	local kvm_available
-
-	if [ -c /dev/kvm ] && eval "$@"; then
-		kvm_available=yes
-	fi
-
-	if [ "$ACCEL" = "kvm" ] && [ "$kvm_available" != "yes" ]; then
+	if [ "$ACCEL" = "kvm" ] && ! kvm_available; then
 		echo "KVM is needed, but not available on this host" >&2
 		return 2
 	fi
 
 	if [ "$ACCEL" ]; then
 		echo $ACCEL
-	elif [ "$kvm_available" = "yes" ]; then
+	elif kvm_available; then
 		echo kvm
 	else
 		echo tcg

  reply	other threads:[~2017-06-29 13:47 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-06-28 20:08 [kvm-unit-tests PATCH 0/5] Minor fixes while getting rid of code duplication Radim Krčmář
2017-06-28 20:08 ` [kvm-unit-tests PATCH 1/5] arch/run: simplify copy-paste of the command line Radim Krčmář
2017-06-28 20:08 ` [kvm-unit-tests PATCH 2/5] run_tests: add summary to powerpc and s390 Radim Krčmář
2017-06-29  5:30   ` Thomas Huth
2017-06-29 12:53     ` Radim Krčmář
2017-06-28 20:08 ` [kvm-unit-tests PATCH 3/5] s390+powerpc/run: improve output handling Radim Krčmář
2017-06-28 20:08 ` [kvm-unit-tests PATCH 4/5] scripts/arch-run: fix qemu binary search failure path Radim Krčmář
2017-06-28 20:46   ` Paolo Bonzini
2017-06-29 12:32     ` Radim Krčmář
2017-06-28 20:08 ` [kvm-unit-tests PATCH 5/5] arch/run: unify accelerator detection Radim Krčmář
2017-06-28 20:48   ` Paolo Bonzini
2017-06-29 12:45     ` Radim Krčmář
2017-06-29 13:46       ` Radim Krčmář [this message]
2017-06-29 13:49         ` Paolo Bonzini
2017-06-29 14:28   ` [kvm-unit-tests PATCH v2 " Radim Krčmář
2017-06-29 14:32   ` [kvm-unit-tests PATCH v2 6/5] x86/run: support accelerator controls Radim Krčmář
2017-06-30 10:42     ` Paolo Bonzini
2017-06-29 11:41 ` [kvm-unit-tests PATCH 0/5] Minor fixes while getting rid of code duplication Andrew Jones

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170629134657.GA14932@potion \
    --to=rkrcmar@redhat.com \
    --cc=david@redhat.com \
    --cc=drjones@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=lvivier@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=thuth@redhat.com \
    --subject='Re: [kvm-unit-tests PATCH 5/5] arch/run: unify accelerator detection' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.