Hello, This series adds Aarch64 support and makes some minor tweaks.
The first two patches of this series add Aarch64 support to libhugetlbfs. (Starting from 3.11-rc1, the Linux Kernel supports HugeTLB and THP for ARM64).
Some general changes are also made: PROT_NONE is added to the mprotect unit test, and the linkhuge_rw test is enabled for 64 bit where there aren't any custom ldscripts.
The final patch clears up the superfluous ARM ld.hugetlbfs HTLB_LINK logic.
Any comments would be appreciated.
Cheers, -- Steve
Steve Capper (5): Aarch64 support. Aarch64 unit test fixes. Add PROT_NONE to the mprotect test. Add linkhuge_rw test to 64 bit && !CUSTOM_LDSCIPTS Cleanup ARM ld.hugetlbfs HTLB_LINK logic
Makefile | 7 +++++++ ld.hugetlbfs | 7 +------ sys-aarch64elf.S | 34 ++++++++++++++++++++++++++++++++++ tests/Makefile | 2 +- tests/icache-hygiene.c | 7 ++++--- tests/mprotect.c | 6 ++++++ tests/mremap-expand-slice-collision.c | 2 +- 7 files changed, 54 insertions(+), 11 deletions(-) create mode 100644 sys-aarch64elf.S
This patch adds support for Aarch64.
As with ARMv7, We do not add the xBT/xBDT style linker scripts as these have been deprecated in favour of adjusting the page sizes via command line parameter to ld.
Signed-off-by: Steve Capper steve.capper@linaro.org --- Makefile | 7 +++++++ ld.hugetlbfs | 2 +- sys-aarch64elf.S | 34 ++++++++++++++++++++++++++++++++++ 3 files changed, 42 insertions(+), 1 deletion(-) create mode 100644 sys-aarch64elf.S
diff --git a/Makefile b/Makefile index 48205af..0e61701 100644 --- a/Makefile +++ b/Makefile @@ -57,6 +57,12 @@ TMPLIB32 = lib ELF32 += armelf_linux_eabi CUSTOM_LDSCRIPTS = no else +ifeq ($(ARCH),aarch64) +CC64 = gcc +ELF64 = aarch64elf +TMPLIB64 = lib64 +CUSTOM_LDSCRIPTS = no +else ifeq ($(ARCH),i386) CC32 = gcc ELF32 = elf_i386 @@ -100,6 +106,7 @@ endif endif endif endif +endif
ifdef CC32 OBJDIRS += obj32 diff --git a/ld.hugetlbfs b/ld.hugetlbfs index d6d12c4..5128aa2 100755 --- a/ld.hugetlbfs +++ b/ld.hugetlbfs @@ -91,7 +91,7 @@ case "$EMU" in elf32ppclinux|elf64ppc) HPAGE_SIZE=$((16*$MB)) SLICE_SIZE=$((256*$MB)) ;; elf_i386|elf_x86_64) HPAGE_SIZE=$((4*$MB)) SLICE_SIZE=$HPAGE_SIZE ;; elf_s390|elf64_s390) HPAGE_SIZE=$((1*$MB)) SLICE_SIZE=$HPAGE_SIZE ;; -armelf_linux_eabi) HPAGE_SIZE=$((2*$MB)) SLICE_SIZE=$HPAGE_SIZE ;; +armelf_linux_eabi|aarch64elf) HPAGE_SIZE=$((2*MB)) SLICE_SIZE=$HPAGE_SIZE ;; esac
if [ "$HTLB_ALIGN" == "slice" ]; then diff --git a/sys-aarch64elf.S b/sys-aarch64elf.S new file mode 100644 index 0000000..54799d3 --- /dev/null +++ b/sys-aarch64elf.S @@ -0,0 +1,34 @@ +/* + * libhugetlbfs - Easy use of Linux hugepages + * Copyright (C) 2013 Linaro Ltd. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public License + * version 2.1 as published by the Free Software Foundation. + * + * This library is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + + .text + + .globl direct_syscall + + +direct_syscall: + uxtw x8, w0 + mov x0, x1 + mov x1, x2 + mov x2, x3 + mov x3, x4 + mov x4, x5 + mov x5, x6 + mov x6, x7 + svc 0x0 + ret
On Aarch64, zero bytes are illegal instructions, this is added to the icache-hygiene test.
In mremap-expand-slice-collision, if __LP64__ is defined then mappings are attempted at 1TB boundaries which are outside the allowable mmap region for Aarch64. For __aarch64__ we change this mapping back to 256MB slices.
Signed-off-by: Steve Capper steve.capper@linaro.org --- tests/icache-hygiene.c | 7 ++++--- tests/mremap-expand-slice-collision.c | 2 +- 2 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/tests/icache-hygiene.c b/tests/icache-hygiene.c index 51792b3..876ce10 100644 --- a/tests/icache-hygiene.c +++ b/tests/icache-hygiene.c @@ -54,7 +54,7 @@ static void cacheflush(void *p) { #if defined(__powerpc__) asm volatile("dcbst 0,%0; sync; icbi 0,%0; isync" : : "r"(p)); -#elif defined(__arm__) +#elif defined(__arm__) || defined(__aarch64__) __clear_cache(p, p + COPY_SIZE); #endif } @@ -87,8 +87,9 @@ static void *sig_expected; static void sig_handler(int signum, siginfo_t *si, void *uc) { #if defined(__powerpc__) || defined(__powerpc64__) || defined(__ia64__) || \ - defined(__s390__) || defined(__s390x__) || defined(__sparc__) - /* On powerpc and ia64 and s390, 0 bytes are an illegal + defined(__s390__) || defined(__s390x__) || defined(__sparc__) || \ + defined(__aarch64__) + /* On powerpc, ia64, s390 and Aarch64, 0 bytes are an illegal * instruction, so, if the icache is cleared properly, we SIGILL * as soon as we jump into the cleared page */ if (signum == SIGILL) { diff --git a/tests/mremap-expand-slice-collision.c b/tests/mremap-expand-slice-collision.c index c25f4c6..853f3c3 100644 --- a/tests/mremap-expand-slice-collision.c +++ b/tests/mremap-expand-slice-collision.c @@ -38,7 +38,7 @@ void init_slice_boundary(int fd) unsigned long slice_size; void *p1, *p2, *heap; int slices_ok, i, rc; -#ifdef __LP64__ +#if defined(__LP64__) && !defined(__aarch64__) /* powerpc: 1TB slices starting at 1 TB */ slice_boundary = 0x10000000000; slice_size = 0x10000000000;
The mprotect unit test checks PROT_READ and PROT_READ | PROT_WRITE protections. We recently found that PROT_NONE wasn't properly supported in an early version of our huge page kernel code.
This patch adds PROT_NONE tests to mprotect. The expected behaviour is that neither reads nor writes should succeed.
Signed-off-by: Steve Capper steve.capper@linaro.org --- tests/mprotect.c | 6 ++++++ 1 file changed, 6 insertions(+)
diff --git a/tests/mprotect.c b/tests/mprotect.c index aa4673e..db6a662 100644 --- a/tests/mprotect.c +++ b/tests/mprotect.c @@ -213,5 +213,11 @@ int main(int argc, char *argv[]) test_mprotect(fd, "RW->R 1/2", 2*hpage_size, PROT_READ|PROT_WRITE, hpage_size, PROT_READ);
+ /* PROT_NONE tests */ + test_mprotect(fd, "NONE->R", hpage_size, PROT_NONE, + hpage_size, PROT_READ); + test_mprotect(fd, "NONE->RW", hpage_size, PROT_NONE, + hpage_size, PROT_READ|PROT_WRITE); + PASS(); }
If one compiles 64 bit with CUSTOM_LDSCRIPTS==no, then the linkhuge_rw test is not compiled even though the logic to build it exists. For 32 bit targets these tests are compiled.
This patch adds $(HUGELINK_RW_TESTS) to the set of tests that are compiled for 64 bit in this case.
Signed-off-by: Steve Capper steve.capper@linaro.org --- tests/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/Makefile b/tests/Makefile index 231e3b0..9140e72 100644 --- a/tests/Makefile +++ b/tests/Makefile @@ -54,7 +54,7 @@ ifeq ($(CUSTOM_LDSCRIPTS),yes) TESTS += $(LDSCRIPT_TESTS) $(HUGELINK_TESTS) $(HUGELINK_TESTS:%=xB.%) \ $(HUGELINK_TESTS:%=xBDT.%) else -TESTS += $(LDSCRIPT_TESTS) $(HUGELINK_TESTS) +TESTS += $(LDSCRIPT_TESTS) $(HUGELINK_TESTS) $(HUGELINK_RW_TESTS) endif
endif
When ld.hugetlbfs is executed with --hugetlbfs-link, there is code to check for the ARM platform and warn that this is not supported.
There is also code to check for CUSTOM_LDSCRIPTS being false and give a similar warning.
This patch removes the ARM check as the CUSTOM_LDSCRIPTS check will catch this.
Signed-off-by: Steve Capper steve.capper@linaro.org --- ld.hugetlbfs | 5 ----- 1 file changed, 5 deletions(-)
diff --git a/ld.hugetlbfs b/ld.hugetlbfs index 5128aa2..ba9e00a 100755 --- a/ld.hugetlbfs +++ b/ld.hugetlbfs @@ -79,11 +79,6 @@ if [ -n "$HTLB_LINK" ]; then HTLB_ALIGN="" # --hugetlbfs-link overrides --hugetlbfs-align LDSCRIPT="$EMU.x$HTLB_LINK" HTLBOPTS="-T${HUGETLB_LDSCRIPT_PATH}/${LDSCRIPT}" - - if [ "$EMU" == "armelf_linux_eabi" ]; then - echo "Please use --hugetlbfs-align when targeting ARM." - exit -1 - fi fi
MB=$((1024*1024))
On Tue, Sep 10, 2013 at 02:11:27PM +0100, Steve Capper wrote:
Hello, This series adds Aarch64 support and makes some minor tweaks.
The first two patches of this series add Aarch64 support to libhugetlbfs. (Starting from 3.11-rc1, the Linux Kernel supports HugeTLB and THP for ARM64).
Some general changes are also made: PROT_NONE is added to the mprotect unit test, and the linkhuge_rw test is enabled for 64 bit where there aren't any custom ldscripts.
The final patch clears up the superfluous ARM ld.hugetlbfs HTLB_LINK logic.
Any comments would be appreciated.
Cheers,
Steve
Hi guys, This is just a ping on this series. Do these patches look ok for inclusion into libhugetlbfs? Should I resend the series?
Thanks,
On 2013-10-18 04:50, Steve Capper wrote:
On Tue, Sep 10, 2013 at 02:11:27PM +0100, Steve Capper wrote:
Hello, This series adds Aarch64 support and makes some minor tweaks.
The first two patches of this series add Aarch64 support to libhugetlbfs. (Starting from 3.11-rc1, the Linux Kernel supports HugeTLB and THP for ARM64).
Some general changes are also made: PROT_NONE is added to the mprotect unit test, and the linkhuge_rw test is enabled for 64 bit where there aren't any custom ldscripts.
The final patch clears up the superfluous ARM ld.hugetlbfs HTLB_LINK logic.
Any comments would be appreciated.
Cheers,
Steve
Hi guys, This is just a ping on this series. Do these patches look ok for inclusion into libhugetlbfs? Should I resend the series?
Thanks,
Steve
Steve Capper (5): Aarch64 support. Aarch64 unit test fixes. Add PROT_NONE to the mprotect test. Add linkhuge_rw test to 64 bit && !CUSTOM_LDSCIPTS Cleanup ARM ld.hugetlbfs HTLB_LINK logic
Makefile | 7 +++++++ ld.hugetlbfs | 7 +------ sys-aarch64elf.S | 34 ++++++++++++++++++++++++++++++++++ tests/Makefile | 2 +- tests/icache-hygiene.c | 7 ++++--- tests/mprotect.c | 6 ++++++ tests/mremap-expand-slice-collision.c | 2 +- 7 files changed, 54 insertions(+), 11 deletions(-) create mode 100644 sys-aarch64elf.S
-- 1.8.1.4
Steve,
The patches are looking good to me (sorry they fell off my radar). I will look to merge them this weekend and cut a new libhugetlbfs release.
Thanks, Eric
On Tue, 10 Sep 2013, Steve Capper wrote:
Hello, This series adds Aarch64 support and makes some minor tweaks.
The first two patches of this series add Aarch64 support to libhugetlbfs. (Starting from 3.11-rc1, the Linux Kernel supports HugeTLB and THP for ARM64).
Some general changes are also made: PROT_NONE is added to the mprotect unit test, and the linkhuge_rw test is enabled for 64 bit where there aren't any custom ldscripts.
The final patch clears up the superfluous ARM ld.hugetlbfs HTLB_LINK logic.
Any comments would be appreciated.
Cheers,
Steve
Sorry it didn't happend this weekend, but these are now applied. I will try and cut a new libhugetlbfs tonight as well.
Thanks, Eric
No probs, thank you Eric.
Cheers, -- Steve
-
Eric B Munson -
Munson, Eric
-
Steve Capper