[OE-core] [PATCH v2 8/9] gcc: backport patch affecting Linux kernel builds
Peter A. Bigot
pab at pabigot.com
Thu Aug 14 19:05:58 UTC 2014
A long-standing bug in gcc turns out to cause problems with unpatched
Linux versions due to improved optimization enabled by gcc 4.9. The
upstream fix missed the gcc-4.9.1 cut-off. It's also been applied
upstream to the 4.8 branch so is being added for OE's 4.8 as well.
Signed-off-by: Peter A. Bigot <pab at pabigot.com>
---
meta/recipes-devtools/gcc/gcc-4.8.inc | 1 +
.../gcc-4.8/0052-PR-rtl-optimization-61801.patch | 36 ++++++++++++++++++++++
meta/recipes-devtools/gcc/gcc-4.9.inc | 1 +
.../gcc-4.9/0055-PR-rtl-optimization-61801.patch | 36 ++++++++++++++++++++++
4 files changed, 74 insertions(+)
create mode 100644 meta/recipes-devtools/gcc/gcc-4.8/0052-PR-rtl-optimization-61801.patch
create mode 100644 meta/recipes-devtools/gcc/gcc-4.9/0055-PR-rtl-optimization-61801.patch
diff --git a/meta/recipes-devtools/gcc/gcc-4.8.inc b/meta/recipes-devtools/gcc/gcc-4.8.inc
index dbe1ba6..18a6692 100644
--- a/meta/recipes-devtools/gcc/gcc-4.8.inc
+++ b/meta/recipes-devtools/gcc/gcc-4.8.inc
@@ -72,6 +72,7 @@ SRC_URI = "\
file://0048-PR58854_fix_arm_apcs_epilogue.patch \
file://0049-Enable-SPE-AltiVec-generation-on-powepc-linux-target.patch \
file://0050-PR-target-58595.patch \
+ file://0052-PR-rtl-optimization-61801.patch \
"
SRC_URI[md5sum] = "a3d7d63b9cb6b6ea049469a0c4a43c9d"
SRC_URI[sha256sum] = "09dc2276c73424bbbfda1dbddc62bbbf900c9f185acf7f3e1d773ce2d7e3cdc8"
diff --git a/meta/recipes-devtools/gcc/gcc-4.8/0052-PR-rtl-optimization-61801.patch b/meta/recipes-devtools/gcc/gcc-4.8/0052-PR-rtl-optimization-61801.patch
new file mode 100644
index 0000000..300b204
--- /dev/null
+++ b/meta/recipes-devtools/gcc/gcc-4.8/0052-PR-rtl-optimization-61801.patch
@@ -0,0 +1,36 @@
+From 8dfe30973ab03f0480f1e249f320e78f210dc230 Mon Sep 17 00:00:00 2001
+From: rguenth <rguenth at 138bc75d-0d04-0410-961f-82ee72b054a4>
+Date: Thu, 17 Jul 2014 07:49:44 +0000
+Subject: [PATCH] 2014-07-17 Richard Biener <rguenther at suse.de>
+
+ PR rtl-optimization/61801
+ * sched-deps.c (sched_analyze_2): For ASM_OPERANDS and
+ ASM_INPUT don't set reg_pending_barrier if it appears in a
+ debug-insn.
+
+
+git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/branches/gcc-4_8-branch@212740 138bc75d-0d04-0410-961f-82ee72b054a4
+
+Upstream-Status: Backport [https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61801]
+Signed-off-by: Peter A. Bigot <pab at pabigot.com>
+
+---
+ gcc/sched-deps.c | 3 ++-
+
+diff --git a/gcc/sched-deps.c b/gcc/sched-deps.c
+index 07857f2..4ac2542 100644
+--- a/gcc/sched-deps.c
++++ b/gcc/sched-deps.c
+@@ -2744,7 +2744,8 @@ sched_analyze_2 (struct deps_desc *deps, rtx x, rtx insn)
+ Consider for instance a volatile asm that changes the fpu rounding
+ mode. An insn should not be moved across this even if it only uses
+ pseudo-regs because it might give an incorrectly rounded result. */
+- if (code != ASM_OPERANDS || MEM_VOLATILE_P (x))
++ if ((code != ASM_OPERANDS || MEM_VOLATILE_P (x))
++ && !DEBUG_INSN_P (insn))
+ reg_pending_barrier = TRUE_BARRIER;
+
+ /* For all ASM_OPERANDS, we must traverse the vector of input operands.
+--
+1.8.5.5
+
diff --git a/meta/recipes-devtools/gcc/gcc-4.9.inc b/meta/recipes-devtools/gcc/gcc-4.9.inc
index bdea3405e..25778bd 100644
--- a/meta/recipes-devtools/gcc/gcc-4.9.inc
+++ b/meta/recipes-devtools/gcc/gcc-4.9.inc
@@ -69,6 +69,7 @@ SRC_URI = "\
file://0051-eabispe.patch \
file://0053-gcc-fix-segfault-from-calling-free-on-non-malloc-d-a.patch \
file://0054-gcc-Makefile.in-fix-parallel-building-failure.patch \
+ file://0055-PR-rtl-optimization-61801.patch \
"
SRC_URI[md5sum] = "fddf71348546af523353bd43d34919c1"
SRC_URI[sha256sum] = "d334781a124ada6f38e63b545e2a3b8c2183049515a1abab6d513f109f1d717e"
diff --git a/meta/recipes-devtools/gcc/gcc-4.9/0055-PR-rtl-optimization-61801.patch b/meta/recipes-devtools/gcc/gcc-4.9/0055-PR-rtl-optimization-61801.patch
new file mode 100644
index 0000000..b27abde
--- /dev/null
+++ b/meta/recipes-devtools/gcc/gcc-4.9/0055-PR-rtl-optimization-61801.patch
@@ -0,0 +1,36 @@
+From 556537c4ad0df4cbebb74197bb2bdea75cf5dd35 Mon Sep 17 00:00:00 2001
+From: rguenth <rguenth at 138bc75d-0d04-0410-961f-82ee72b054a4>
+Date: Thu, 17 Jul 2014 07:48:49 +0000
+Subject: [PATCH] 2014-07-17 Richard Biener <rguenther at suse.de>
+
+ PR rtl-optimization/61801
+ * sched-deps.c (sched_analyze_2): For ASM_OPERANDS and
+ ASM_INPUT don't set reg_pending_barrier if it appears in a
+ debug-insn.
+
+
+git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/branches/gcc-4_9-branch@212739 138bc75d-0d04-0410-961f-82ee72b054a4
+
+Upstream-Status: Backport [https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61801]
+Signed-off-by: Peter A. Bigot <pab at pabigot.com>
+
+---
+ gcc/sched-deps.c | 3 ++-
+
+diff --git a/gcc/sched-deps.c b/gcc/sched-deps.c
+index efc4223..df29bd3 100644
+--- a/gcc/sched-deps.c
++++ b/gcc/sched-deps.c
+@@ -2750,7 +2750,8 @@ sched_analyze_2 (struct deps_desc *deps, rtx x, rtx insn)
+ Consider for instance a volatile asm that changes the fpu rounding
+ mode. An insn should not be moved across this even if it only uses
+ pseudo-regs because it might give an incorrectly rounded result. */
+- if (code != ASM_OPERANDS || MEM_VOLATILE_P (x))
++ if ((code != ASM_OPERANDS || MEM_VOLATILE_P (x))
++ && !DEBUG_INSN_P (insn))
+ reg_pending_barrier = TRUE_BARRIER;
+
+ /* For all ASM_OPERANDS, we must traverse the vector of input operands.
+--
+1.8.5.5
+
--
1.8.5.5
More information about the Openembedded-core
mailing list