All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jeff Dike <jdike@addtoit.com>
To: Andrew Morton <akpm@osdl.org>
Cc: LKML <linux-kernel@vger.kernel.org>,
	uml-devel <user-mode-linux-devel@lists.sourceforge.net>,
	Karol Swietlicki <magotari@gmail.com>
Subject: [uml-devel] [PATCH 3/6] UML - Improve detection of host cmov
Date: Fri, 2 Nov 2007 11:40:14 -0400	[thread overview]
Message-ID: <20071102154014.GA10263@c2.user-mode-linux.org> (raw)

From: Karol Swietlicki <magotari@gmail.com>

This patch introduces a new way of checking for the cmov instruction.
I use signal handling instead of reading /proc/cpuinfo.

[ jdike - Fiddled the asm to make it obvious that it didn't mess with
any in-use registers and made test_for_host_cmov void ]

Signed-off-by: Karol Swietlicki <magotari@gmail.com>
Signed-off-by: Jeff Dike <jdike@linux.intel.com>
---
 arch/um/sys-i386/bugs.c |   41 ++++++++++++++++++++++++++++++++---------
 1 file changed, 32 insertions(+), 9 deletions(-)

Index: linux-2.6.22/arch/um/sys-i386/bugs.c
===================================================================
--- linux-2.6.22.orig/arch/um/sys-i386/bugs.c	2007-10-30 12:31:35.000000000 -0400
+++ linux-2.6.22/arch/um/sys-i386/bugs.c	2007-10-31 11:24:45.000000000 -0400
@@ -10,11 +10,42 @@
 #include "os.h"
 #include "task.h"
 #include "user.h"
+#include "sysdep/archsetjmp.h"
 
 #define MAXTOKEN 64
 
 /* Set during early boot */
 int host_has_cmov = 1;
+static jmp_buf cmov_test_return;
+
+static void cmov_sigill_test_handler(int sig)
+{
+	host_has_cmov = 0;
+	longjmp(cmov_test_return, 1);
+}
+
+static void test_for_host_cmov(void)
+{
+	struct sigaction old, new;
+
+	printk(UM_KERN_INFO "Checking for host processor cmov support...");
+	new.sa_handler = cmov_sigill_test_handler;
+
+	/* Make sure that SIGILL is enabled after the handler longjmps back */
+	new.sa_flags = SA_NODEFER;
+	sigemptyset(&new.sa_mask);
+	sigaction(SIGILL, &new, &old);
+
+	if (setjmp(cmov_test_return) == 0) {
+		unsigned long foo = 0;
+		__asm__ __volatile__("cmovz %0, %1" : "=r" (foo) : "0" (foo));
+		printk("Yes\n");
+	}
+	else
+		printk("No\n");
+
+	sigaction(SIGILL, &old, &new);
+}
 
 static char token(int fd, char *buf, int len, char stop)
 {
@@ -153,15 +184,7 @@ void arch_init_thread(void)
 
 void arch_check_bugs(void)
 {
-	int have_it;
-
-	if (os_access("/proc/cpuinfo", OS_ACC_R_OK) < 0) {
-		printk(UM_KERN_ERR "/proc/cpuinfo not available - skipping CPU "
-		       "capability checks\n");
-		return;
-	}
-	if (check_cpu_flag("cmov", &have_it))
-		host_has_cmov = have_it;
+	test_for_host_cmov();
 }
 
 int arch_handle_signal(int sig, struct uml_pt_regs *regs)

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

WARNING: multiple messages have this Message-ID (diff)
From: Jeff Dike <jdike@addtoit.com>
To: Andrew Morton <akpm@osdl.org>
Cc: Karol Swietlicki <magotari@gmail.com>,
	LKML <linux-kernel@vger.kernel.org>,
	uml-devel <user-mode-linux-devel@lists.sourceforge.net>
Subject: [PATCH 3/6] UML - Improve detection of host cmov
Date: Fri, 2 Nov 2007 11:40:14 -0400	[thread overview]
Message-ID: <20071102154014.GA10263@c2.user-mode-linux.org> (raw)

From: Karol Swietlicki <magotari@gmail.com>

This patch introduces a new way of checking for the cmov instruction.
I use signal handling instead of reading /proc/cpuinfo.

[ jdike - Fiddled the asm to make it obvious that it didn't mess with
any in-use registers and made test_for_host_cmov void ]

Signed-off-by: Karol Swietlicki <magotari@gmail.com>
Signed-off-by: Jeff Dike <jdike@linux.intel.com>
---
 arch/um/sys-i386/bugs.c |   41 ++++++++++++++++++++++++++++++++---------
 1 file changed, 32 insertions(+), 9 deletions(-)

Index: linux-2.6.22/arch/um/sys-i386/bugs.c
===================================================================
--- linux-2.6.22.orig/arch/um/sys-i386/bugs.c	2007-10-30 12:31:35.000000000 -0400
+++ linux-2.6.22/arch/um/sys-i386/bugs.c	2007-10-31 11:24:45.000000000 -0400
@@ -10,11 +10,42 @@
 #include "os.h"
 #include "task.h"
 #include "user.h"
+#include "sysdep/archsetjmp.h"
 
 #define MAXTOKEN 64
 
 /* Set during early boot */
 int host_has_cmov = 1;
+static jmp_buf cmov_test_return;
+
+static void cmov_sigill_test_handler(int sig)
+{
+	host_has_cmov = 0;
+	longjmp(cmov_test_return, 1);
+}
+
+static void test_for_host_cmov(void)
+{
+	struct sigaction old, new;
+
+	printk(UM_KERN_INFO "Checking for host processor cmov support...");
+	new.sa_handler = cmov_sigill_test_handler;
+
+	/* Make sure that SIGILL is enabled after the handler longjmps back */
+	new.sa_flags = SA_NODEFER;
+	sigemptyset(&new.sa_mask);
+	sigaction(SIGILL, &new, &old);
+
+	if (setjmp(cmov_test_return) == 0) {
+		unsigned long foo = 0;
+		__asm__ __volatile__("cmovz %0, %1" : "=r" (foo) : "0" (foo));
+		printk("Yes\n");
+	}
+	else
+		printk("No\n");
+
+	sigaction(SIGILL, &old, &new);
+}
 
 static char token(int fd, char *buf, int len, char stop)
 {
@@ -153,15 +184,7 @@ void arch_init_thread(void)
 
 void arch_check_bugs(void)
 {
-	int have_it;
-
-	if (os_access("/proc/cpuinfo", OS_ACC_R_OK) < 0) {
-		printk(UM_KERN_ERR "/proc/cpuinfo not available - skipping CPU "
-		       "capability checks\n");
-		return;
-	}
-	if (check_cpu_flag("cmov", &have_it))
-		host_has_cmov = have_it;
+	test_for_host_cmov();
 }
 
 int arch_handle_signal(int sig, struct uml_pt_regs *regs)

             reply	other threads:[~2007-11-02 15:40 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-11-02 15:40 Jeff Dike [this message]
2007-11-02 15:40 ` [PATCH 3/6] UML - Improve detection of host cmov Jeff Dike

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=20071102154014.GA10263@c2.user-mode-linux.org \
    --to=jdike@addtoit.com \
    --cc=akpm@osdl.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=magotari@gmail.com \
    --cc=user-mode-linux-devel@lists.sourceforge.net \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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.