Add patch to fix loop in VDSO (rhbz 1178975)

kernel.spec

@@ -614,6 +614,9 @@ Patch26092: xhci-Add-broken-streams-quirk-for-Fresco-Logic-FL100.patch
 Patch26093: uas-Add-US_FL_NO_ATA_1X-for-Seagate-devices-with-usb.patch
 Patch26094: uas-Add-US_FL_NO_REPORT_OPCODES-for-JMicron-JMS566-w.patch
 
+#rhbz 1178975
+Patch26125: x86-vdso-Use-asm-volatile-in-__getcpu.patch
+
 # git clone ssh://git.fedorahosted.org/git/kernel-arm64.git, git diff master...devel
 Patch30000: kernel-arm64.patch
 
@@ -1336,6 +1339,9 @@ ApplyPatch xhci-Add-broken-streams-quirk-for-Fresco-Logic-FL100.patch
 ApplyPatch uas-Add-US_FL_NO_ATA_1X-for-Seagate-devices-with-usb.patch
 ApplyPatch uas-Add-US_FL_NO_REPORT_OPCODES-for-JMicron-JMS566-w.patch
 
+#rhbz 1178975
+ApplyPatch x86-vdso-Use-asm-volatile-in-__getcpu.patch
+
 %if 0%{?aarch64patches}
 ApplyPatch kernel-arm64.patch
 %ifnarch aarch64 # this is stupid, but i want to notice before secondary koji does.
@@ -2202,6 +2208,9 @@ fi
 #                                    ||----w |
 #                                    ||     ||
 %changelog
+* Mon Jan 12 2015 Josh Boyer <jwboyer@fedoraproject.org>
+- Add patch to fix loop in VDSO (rhbz 1178975)
+
 * Fri Jan 09 2015 Josh Boyer <jwboyer@fedoraproject.org> - 3.19.0-0.rc3.git2.1
 - Linux v3.19-rc3-69-g11c8f01b423b
 

x86-vdso-Use-asm-volatile-in-__getcpu.patch

@@ -0,0 +1,55 @@
+From 1ddf0b1b11aa8a90cef6706e935fc31c75c406ba Mon Sep 17 00:00:00 2001
+From: Andy Lutomirski <luto@amacapital.net>
+Date: Sun, 21 Dec 2014 08:57:46 -0800
+Subject: x86, vdso: Use asm volatile in __getcpu
+
+In Linux 3.18 and below, GCC hoists the lsl instructions in the
+pvclock code all the way to the beginning of __vdso_clock_gettime,
+slowing the non-paravirt case significantly.  For unknown reasons,
+presumably related to the removal of a branch, the performance issue
+is gone as of
+
+e76b027e6408 x86,vdso: Use LSL unconditionally for vgetcpu
+
+but I don't trust GCC enough to expect the problem to stay fixed.
+
+There should be no correctness issue, because the __getcpu calls in
+__vdso_vlock_gettime were never necessary in the first place.
+
+Note to stable maintainers: In 3.18 and below, depending on
+configuration, gcc 4.9.2 generates code like this:
+
+     9c3:       44 0f 03 e8             lsl    %ax,%r13d
+     9c7:       45 89 eb                mov    %r13d,%r11d
+     9ca:       0f 03 d8                lsl    %ax,%ebx
+
+This patch won't apply as is to any released kernel, but I'll send a
+trivial backported version if needed.
+
+Fixes: 51c19b4f5927 x86: vdso: pvclock gettime support
+Cc: stable@vger.kernel.org # 3.8+
+Cc: Marcelo Tosatti <mtosatti@redhat.com>
+Acked-by: Paolo Bonzini <pbonzini@redhat.com>
+Signed-off-by: Andy Lutomirski <luto@amacapital.net>
+
+diff --git a/arch/x86/include/asm/vgtod.h b/arch/x86/include/asm/vgtod.h
+index e7e9682..f556c48 100644
+--- a/arch/x86/include/asm/vgtod.h
++++ b/arch/x86/include/asm/vgtod.h
+@@ -80,9 +80,11 @@ static inline unsigned int __getcpu(void)
+ 
+ 	/*
+ 	 * Load per CPU data from GDT.  LSL is faster than RDTSCP and
+-	 * works on all CPUs.
++	 * works on all CPUs.  This is volatile so that it orders
++	 * correctly wrt barrier() and to keep gcc from cleverly
++	 * hoisting it out of the calling function.
+ 	 */
+-	asm("lsl %1,%0" : "=r" (p) : "r" (__PER_CPU_SEG));
++	asm volatile ("lsl %1,%0" : "=r" (p) : "r" (__PER_CPU_SEG));
+ 
+ 	return p;
+ }
+-- 
+cgit v0.10.2
+