diff mbox

[Xen-devel,v2,for-4.8] tools/libacpi: Fix compilation when cross building the tools

Message ID 1480338816-29434-1-git-send-email-julien.grall@arm.com
State New
Headers show

Commit Message

Julien Grall Nov. 28, 2016, 1:13 p.m. UTC
The tools (such as mk_dsdt) can be cross-built when it may not be
desirable to build them on the target.

The commit c4ac1077 "libxl/arm: Generate static ACPI DSDT table"
introduced support of ARM64 in mk_dsdt but also break cross-building
tools because the ACPI tables are not correct.

While mk_dsdt should generate ACPI table for the target architecture, it
currently generates the one for the host. This is because the source
code contains reference to the host architecture (__aarch64__,
__x86_64__, __i386__) when it should be the target architecture.

Replace all __aarch64__, __x86_64__, __i386__ by the corresponding
CONFIG_*.

Also expose the CONFIG_* to the source code as the currently only
exposed to the Makefile.

Reported-by: Andrii Anisov <andrii.anisov@gmail.com>
Suggested-by: Wei Liu <wei.liu2@citrix.com>
Signed-off-by: Julien Grall <julien.grall@arm.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

---
    Changes in v2:
        - Use defined(CONFIG_*) rather than CONFIG_*
        - Use #if defined #elif defined construction
        - Directly use MKDSDT_CFLAGS-y
        - Add Andrew's reviewed-by

    This was reported on the ML recently (see [1]) and affects only Xen
    4.8. Without this patch, cross-building the tools will not work.

    I think this patch is quite important for embedded users where they
    tend to cross-build the rootfs (for instance using yocto).

    The patch is fairly simple, exposing CONFIG_* to the source code and
    replacing all reference to the host architecture to the
    corresponding target architecture. It could be easy to test all the
    configuration.

    I diffed the generated dsdt and it is the same before and after the
    patches when built natively. I did try cross-build, Andrii could you
    give a try?

    [1] https://lists.xenproject.org/archives/html/xen-devel/2016-11/msg01903.html
---
 tools/libacpi/Makefile  |  5 ++++-
 tools/libacpi/mk_dsdt.c | 14 +++++++-------
 2 files changed, 11 insertions(+), 8 deletions(-)

Comments

Jan Beulich Nov. 28, 2016, 1:30 p.m. UTC | #1
>>> On 28.11.16 at 14:13, <julien.grall@arm.com> wrote:
> The tools (such as mk_dsdt) can be cross-built when it may not be
> desirable to build them on the target.
> 
> The commit c4ac1077 "libxl/arm: Generate static ACPI DSDT table"
> introduced support of ARM64 in mk_dsdt but also break cross-building
> tools because the ACPI tables are not correct.
> 
> While mk_dsdt should generate ACPI table for the target architecture, it
> currently generates the one for the host. This is because the source
> code contains reference to the host architecture (__aarch64__,
> __x86_64__, __i386__) when it should be the target architecture.
> 
> Replace all __aarch64__, __x86_64__, __i386__ by the corresponding
> CONFIG_*.
> 
> Also expose the CONFIG_* to the source code as the currently only
> exposed to the Makefile.
> 
> Reported-by: Andrii Anisov <andrii.anisov@gmail.com>
> Suggested-by: Wei Liu <wei.liu2@citrix.com>
> Signed-off-by: Julien Grall <julien.grall@arm.com>
> Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

Reviewed-by: Jan Beulich <jbeulich@suse.com>

However, ...

> --- a/tools/libacpi/mk_dsdt.c
> +++ b/tools/libacpi/mk_dsdt.c
> @@ -17,9 +17,9 @@
>  #include <getopt.h>
>  #include <stdlib.h>
>  #include <stdbool.h>
> -#if defined(__i386__) || defined(__x86_64__)
> +#if defined(CONFIG_X86)
>  #include <xen/hvm/hvm_info_table.h>
> -#elif defined(__aarch64__)
> +#elif defined(CONFIG_ARM_64)
>  #include <xen/arch-arm.h>
>  #endif

.. for this and at least some of the others I wonder whether from an
abstract pov these shouldn't be CONFIG_ARM. Agreed, it won't
matter as long as there's no use of ACPI on ARM32, hence my R-b
stands either way, but I'd like you to clarify whether it really should
go in this way.

Jan
Julien Grall Nov. 28, 2016, 1:37 p.m. UTC | #2
Hi,

On 28/11/16 13:30, Jan Beulich wrote:
>>>> On 28.11.16 at 14:13, <julien.grall@arm.com> wrote:
>> The tools (such as mk_dsdt) can be cross-built when it may not be
>> desirable to build them on the target.
>>
>> The commit c4ac1077 "libxl/arm: Generate static ACPI DSDT table"
>> introduced support of ARM64 in mk_dsdt but also break cross-building
>> tools because the ACPI tables are not correct.
>>
>> While mk_dsdt should generate ACPI table for the target architecture, it
>> currently generates the one for the host. This is because the source
>> code contains reference to the host architecture (__aarch64__,
>> __x86_64__, __i386__) when it should be the target architecture.
>>
>> Replace all __aarch64__, __x86_64__, __i386__ by the corresponding
>> CONFIG_*.
>>
>> Also expose the CONFIG_* to the source code as the currently only
>> exposed to the Makefile.
>>
>> Reported-by: Andrii Anisov <andrii.anisov@gmail.com>
>> Suggested-by: Wei Liu <wei.liu2@citrix.com>
>> Signed-off-by: Julien Grall <julien.grall@arm.com>
>> Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
>
> Reviewed-by: Jan Beulich <jbeulich@suse.com>
>
> However, ...
>
>> --- a/tools/libacpi/mk_dsdt.c
>> +++ b/tools/libacpi/mk_dsdt.c
>> @@ -17,9 +17,9 @@
>>  #include <getopt.h>
>>  #include <stdlib.h>
>>  #include <stdbool.h>
>> -#if defined(__i386__) || defined(__x86_64__)
>> +#if defined(CONFIG_X86)
>>  #include <xen/hvm/hvm_info_table.h>
>> -#elif defined(__aarch64__)
>> +#elif defined(CONFIG_ARM_64)
>>  #include <xen/arch-arm.h>
>>  #endif
>
> .. for this and at least some of the others I wonder whether from an
> abstract pov these shouldn't be CONFIG_ARM. Agreed, it won't
> matter as long as there's no use of ACPI on ARM32, hence my R-b
> stands either way, but I'd like you to clarify whether it really should
> go in this way.

To answer the question, I am not aware of any plan to have support for 
ACPI on ARM32.

Regardless the answer, this would be a separate patch as CONFIG_ARM_64 
is the correct define to match __aarch64__. Let's not mix improvement 
and bug fix.

Regards,
Jan Beulich Nov. 28, 2016, 1:45 p.m. UTC | #3
>>> On 28.11.16 at 14:37, <julien.grall@arm.com> wrote:
> On 28/11/16 13:30, Jan Beulich wrote:
>> However, ...
>>
>>> --- a/tools/libacpi/mk_dsdt.c
>>> +++ b/tools/libacpi/mk_dsdt.c
>>> @@ -17,9 +17,9 @@
>>>  #include <getopt.h>
>>>  #include <stdlib.h>
>>>  #include <stdbool.h>
>>> -#if defined(__i386__) || defined(__x86_64__)
>>> +#if defined(CONFIG_X86)
>>>  #include <xen/hvm/hvm_info_table.h>
>>> -#elif defined(__aarch64__)
>>> +#elif defined(CONFIG_ARM_64)
>>>  #include <xen/arch-arm.h>
>>>  #endif
>>
>> .. for this and at least some of the others I wonder whether from an
>> abstract pov these shouldn't be CONFIG_ARM. Agreed, it won't
>> matter as long as there's no use of ACPI on ARM32, hence my R-b
>> stands either way, but I'd like you to clarify whether it really should
>> go in this way.
> 
> To answer the question, I am not aware of any plan to have support for 
> ACPI on ARM32.

Except that this wasn't the question.

> Regardless the answer, this would be a separate patch as CONFIG_ARM_64 
> is the correct define to match __aarch64__. Let's not mix improvement 
> and bug fix.

Well, it looks even more suspicious to me now than it did before
that both x86-32 and x86-64 are being taken care of, but only
arm64. But anyway ...

Jan
Wei Liu Nov. 28, 2016, 2:23 p.m. UTC | #4
On Mon, Nov 28, 2016 at 06:30:58AM -0700, Jan Beulich wrote:
> >>> On 28.11.16 at 14:13, <julien.grall@arm.com> wrote:
> > The tools (such as mk_dsdt) can be cross-built when it may not be
> > desirable to build them on the target.
> > 
> > The commit c4ac1077 "libxl/arm: Generate static ACPI DSDT table"
> > introduced support of ARM64 in mk_dsdt but also break cross-building
> > tools because the ACPI tables are not correct.
> > 
> > While mk_dsdt should generate ACPI table for the target architecture, it
> > currently generates the one for the host. This is because the source
> > code contains reference to the host architecture (__aarch64__,
> > __x86_64__, __i386__) when it should be the target architecture.
> > 
> > Replace all __aarch64__, __x86_64__, __i386__ by the corresponding
> > CONFIG_*.
> > 
> > Also expose the CONFIG_* to the source code as the currently only
> > exposed to the Makefile.
> > 
> > Reported-by: Andrii Anisov <andrii.anisov@gmail.com>
> > Suggested-by: Wei Liu <wei.liu2@citrix.com>
> > Signed-off-by: Julien Grall <julien.grall@arm.com>
> > Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
> 
> Reviewed-by: Jan Beulich <jbeulich@suse.com>
> 

Release-acked-by: Wei Liu <wei.liu2@citrix.com>
Andrii Anisov Nov. 28, 2016, 4:44 p.m. UTC | #5
>     I diffed the generated dsdt and it is the same before and after the
>     patches when built natively. I did try cross-build, Andrii could you
>     give a try?
Build passed smoothly.
Non-ACPI guests started well.

Sincerely,
Andrii Anisov.
diff mbox

Patch

diff --git a/tools/libacpi/Makefile b/tools/libacpi/Makefile
index ccc32c9..6d8445d 100644
--- a/tools/libacpi/Makefile
+++ b/tools/libacpi/Makefile
@@ -27,6 +27,9 @@  DSDT_FILES ?= $(C_SRC-y)
 C_SRC = $(addprefix $(ACPI_BUILD_DIR)/, $(DSDT_FILES))
 H_SRC = $(addprefix $(ACPI_BUILD_DIR)/, ssdt_s3.h ssdt_s4.h ssdt_pm.h ssdt_tpm.h)
 
+MKDSDT_CFLAGS-$(CONFIG_ARM_64) = -DCONFIG_ARM_64
+MKDSDT_CFLAGS-$(CONFIG_X86) = -DCONFIG_X86
+
 # Suffix for temporary files.
 #
 # We will also use this suffix to workaround a bug in older iasl
@@ -44,7 +47,7 @@  $(H_SRC): $(ACPI_BUILD_DIR)/%.h: %.asl iasl
 	rm -f $(addprefix $(ACPI_BUILD_DIR)/, $*.aml $*.hex)
  
 $(MK_DSDT): mk_dsdt.c
-	$(HOSTCC) $(HOSTCFLAGS) $(CFLAGS_xeninclude) -D__XEN_TOOLS__ -o $@ mk_dsdt.c
+	$(HOSTCC) $(HOSTCFLAGS) $(MKDSDT_CFLAGS-y) $(CFLAGS_xeninclude) -D__XEN_TOOLS__ -o $@ mk_dsdt.c
 
 $(ACPI_BUILD_DIR)/dsdt_anycpu_qemu_xen.asl: dsdt.asl dsdt_acpi_info.asl $(MK_DSDT)
 	# Remove last bracket
diff --git a/tools/libacpi/mk_dsdt.c b/tools/libacpi/mk_dsdt.c
index 16320a9..760d81b 100644
--- a/tools/libacpi/mk_dsdt.c
+++ b/tools/libacpi/mk_dsdt.c
@@ -17,9 +17,9 @@ 
 #include <getopt.h>
 #include <stdlib.h>
 #include <stdbool.h>
-#if defined(__i386__) || defined(__x86_64__)
+#if defined(CONFIG_X86)
 #include <xen/hvm/hvm_info_table.h>
-#elif defined(__aarch64__)
+#elif defined(CONFIG_ARM_64)
 #include <xen/arch-arm.h>
 #endif
 
@@ -111,9 +111,9 @@  int main(int argc, char **argv)
     unsigned int slot, dev, intx, link, cpu, max_cpus;
     dm_version dm_version = QEMU_XEN_TRADITIONAL;
 
-#if defined(__i386__) || defined(__x86_64__)
+#if defined(CONFIG_X86)
     max_cpus = HVM_MAX_VCPUS;
-#elif defined(__aarch64__)
+#elif defined(CONFIG_ARM_64)
     max_cpus = GUEST_MAX_VCPUS;
 #endif
 
@@ -169,7 +169,7 @@  int main(int argc, char **argv)
     /**** Processor start ****/
     push_block("Scope", "\\_SB");
 
-#if defined(__i386__) || defined(__x86_64__)
+#ifdef CONFIG_X86
     /* MADT checksum */
     stmt("OperationRegion", "MSUM, SystemMemory, \\_SB.MSUA, 1");
     push_block("Field", "MSUM, ByteAcc, NoLock, Preserve");
@@ -193,7 +193,7 @@  int main(int argc, char **argv)
         stmt("Name", "_HID, \"ACPI0007\"");
 
         stmt("Name", "_UID, %d", cpu);
-#if defined(__aarch64__)
+#ifdef CONFIG_ARM_64
         pop_block();
         continue;
 #endif
@@ -235,7 +235,7 @@  int main(int argc, char **argv)
         pop_block();
     }
 
-#if defined(__aarch64__)
+#ifdef CONFIG_ARM_64
     pop_block();
     /**** Processor end ****/
     pop_block();