2014-01-06 10:31:31

by James Hogan

[permalink] [raw]
Subject: Re: [Suggest] arch: metag: compiler: Are they compiler's issues?

On 24/12/13 02:36, Chen Gang wrote:
> Helllo Maintainers:
>
> When compile metag with allmodconfig, I met the 2 kinds of issues, I
> guess they are compiler's issues, please help check, thanks.
>
>
> Issue 1: compiler report the issue, but after check the related kernel
> source code, I can not find any issues.
>
> CC [M] drivers/isdn/hisax/s0box.o
> drivers/isdn/hisax/hscx_irq.c: In function 'hscx_fill_fifo':
> /upstream/linux-next/arch/metag/include/asm/io.h:62: error: 'asm' operand requires impossible reload
> make[3]: *** [drivers/isdn/hisax/s0box.o] Error 1
> make[2]: *** [drivers/isdn/hisax] Error 2
> make[1]: *** [drivers/isdn] Error 2
> make: *** [drivers] Error 2

This was a compiler bug in an old version of the toolchain.
What does metag-linux-gcc -v give you?
It should be "gcc version 4.2.4 (IMG-1.4.0.700)"

> Issue 2: they are about structure variables alignment, I guess, the
> compiler can not notice about the "#pragma pack()" correctly.
>
> CC [M] drivers/staging/lustre/lustre/lov/lov_pack.o
> drivers/staging/lustre/lustre/lov/lov_pack.c: In function 'lov_getstripe':
> drivers/staging/lustre/lustre/lov/lov_pack.c:630: error: duplicate case value
> drivers/staging/lustre/lustre/lov/lov_pack.c:630: error: previously used here
> make[5]: *** [drivers/staging/lustre/lustre/lov/lov_pack.o] Error 1
> make[4]: *** [drivers/staging/lustre/lustre/lov] Error 2
> make[3]: *** [drivers/staging/lustre/lustre] Error 2
> make[2]: *** [drivers/staging/lustre] Error 2
> make[1]: *** [drivers/staging] Error 2
> make: *** [drivers] Error 2
>
> MODPOST 2909 modules
> ERROR: "__compiletime_assert_426" [net/batman-adv/batman-adv.ko] undefined!
> ERROR: "__compiletime_assert_429" [net/batman-adv/batman-adv.ko] undefined!
> ERROR: "__compiletime_assert_425" [net/batman-adv/batman-adv.ko] undefined!
> ERROR: "__compiletime_assert_428" [net/batman-adv/batman-adv.ko] undefined!
> ERROR: "__compiletime_assert_427" [net/batman-adv/batman-adv.ko] undefined!
> ERROR: "__compiletime_assert_430" [net/batman-adv/batman-adv.ko] undefined!

I suspect this is due to bad assumptions in the code. The metag ABI is
unusual in padding the size of structs to a 32bit boundary even if all
members are <32bit. This is actually permitted by the C standard but
it's a bit of a pain. e.g.

struct s {
short x
struct {
short x[3];
} y;
short z;
};

on x86
alignof(s::y) == 2
s::y at offset 2
sizeof(s::y) == 6
s::z at offset 6+2 = 8
sizeof(struct s) == 10

but on metag
alignof(s::y) == 4
s::y at offset 4
sizeof(s::y) == 8 (padding, this is what catches people out)
s::z at offset 4+8 = 12
sizeof(struct s) == 16 (and here too)

Adding packed attribute on outer struct reduces sizeof(struct s) to 12
on metag:
alignof(s::y) == 4
s::y at offset 2 (packed)
sizeof(s::y) == 8 (still padded)
s::z at offset 2+8 = 10
sizeof(struct s) == 12 (packed)

Also reduced to 12 if only inner struct is marked packed:
alignof(s::y) == 2
s::y at offset 2
sizeof(s::y) == 6 (packed)
s::z at offset 2+6 = 8
sizeof(struct s) == 12 (still padded)

Adding packed attribute on both outer and inner struct reduces
sizeof(struct s) to 10 to match x86.

Unfortunately it's years too late to change this ABI, so we're stuck
with it.

>
>
> Making cross-compiler:
>
> git clone git://github.com/img-meta/metag-buildroot.git metag-core-2013.11
> make meta2_defconfig.

That loads the meta2_defconfig. Did you rebuild it? If in doubt you can
always "rm -fr output" and re-make.

Cheers
James

> "cd metag-core-2013.11; cp -r output/host/usr/* /usr/"
>
>
> Welcome any additional suggestions, discussions or completions.
>
> Thanks.
>


2014-01-08 15:01:34

by Chen Gang

[permalink] [raw]
Subject: Re: [Suggest] arch: metag: compiler: Are they compiler's issues?

On 01/06/2014 06:31 PM, James Hogan wrote:
> On 24/12/13 02:36, Chen Gang wrote:
>> Helllo Maintainers:
>>
>> When compile metag with allmodconfig, I met the 2 kinds of issues, I
>> guess they are compiler's issues, please help check, thanks.
>>
>>
>> Issue 1: compiler report the issue, but after check the related kernel
>> source code, I can not find any issues.
>>
>> CC [M] drivers/isdn/hisax/s0box.o
>> drivers/isdn/hisax/hscx_irq.c: In function 'hscx_fill_fifo':
>> /upstream/linux-next/arch/metag/include/asm/io.h:62: error: 'asm' operand requires impossible reload
>> make[3]: *** [drivers/isdn/hisax/s0box.o] Error 1
>> make[2]: *** [drivers/isdn/hisax] Error 2
>> make[1]: *** [drivers/isdn] Error 2
>> make: *** [drivers] Error 2
>
> This was a compiler bug in an old version of the toolchain.
> What does metag-linux-gcc -v give you?
> It should be "gcc version 4.2.4 (IMG-1.4.0.700)"
>

OK, thanks, and the gcc version is below:

[root@gchen ~]# /usr/bin/metag-buildroot-linux-uclibc-gcc -v
Using built-in specs.
Target: metag-buildroot-linux-uclibc
Configured with: /upstream/src/metag/metag-core-2013.11/output/toolchain/gcc-4.2.4/configure --prefix=/upstream/src/metag/metag-core-2013.11/output/host/usr --build=x86_64-unknown-linux-gnu --host=x86_64-unknown-linux-gnu --target=metag-buildroot-linux-uclibc --enable-languages=c --with-sysroot=/upstream/src/metag/metag-core-2013.11/output/host/usr/metag-buildroot-linux-uclibc/sysroot --with-build-time-tools=/upstream/src/metag/metag-core-2013.11/output/host/usr/metag-buildroot-linux-uclibc/bin --disable-__cxa_atexit --enable-target-optspace --disable-libgomp --with-gnu-ld --disable-libssp --disable-multilib --enable-tls --enable-shared --with-gmp=/upstream/src/metag/metag-core-2013.11/output/host/usr --with-mpfr=/upstream/src/metag/metag-core-2013.11/output/host/usr --disable-nls --enable-threads --with-cpu=2.1 --enable-meta-default --disable-symvers
Thread model: posix
gcc version 4.2.4 (IMG-1.4.0.300)


>> Issue 2: they are about structure variables alignment, I guess, the
>> compiler can not notice about the "#pragma pack()" correctly.
>>
>> CC [M] drivers/staging/lustre/lustre/lov/lov_pack.o
>> drivers/staging/lustre/lustre/lov/lov_pack.c: In function 'lov_getstripe':
>> drivers/staging/lustre/lustre/lov/lov_pack.c:630: error: duplicate case value
>> drivers/staging/lustre/lustre/lov/lov_pack.c:630: error: previously used here
>> make[5]: *** [drivers/staging/lustre/lustre/lov/lov_pack.o] Error 1
>> make[4]: *** [drivers/staging/lustre/lustre/lov] Error 2
>> make[3]: *** [drivers/staging/lustre/lustre] Error 2
>> make[2]: *** [drivers/staging/lustre] Error 2
>> make[1]: *** [drivers/staging] Error 2
>> make: *** [drivers] Error 2
>>
>> MODPOST 2909 modules
>> ERROR: "__compiletime_assert_426" [net/batman-adv/batman-adv.ko] undefined!
>> ERROR: "__compiletime_assert_429" [net/batman-adv/batman-adv.ko] undefined!
>> ERROR: "__compiletime_assert_425" [net/batman-adv/batman-adv.ko] undefined!
>> ERROR: "__compiletime_assert_428" [net/batman-adv/batman-adv.ko] undefined!
>> ERROR: "__compiletime_assert_427" [net/batman-adv/batman-adv.ko] undefined!
>> ERROR: "__compiletime_assert_430" [net/batman-adv/batman-adv.ko] undefined!
>
> I suspect this is due to bad assumptions in the code. The metag ABI is
> unusual in padding the size of structs to a 32bit boundary even if all
> members are <32bit. This is actually permitted by the C standard but
> it's a bit of a pain. e.g.
>
> struct s {
> short x
> struct {
> short x[3];
> } y;
> short z;
> };
>
> on x86
> alignof(s::y) == 2
> s::y at offset 2
> sizeof(s::y) == 6
> s::z at offset 6+2 = 8
> sizeof(struct s) == 10
>
> but on metag
> alignof(s::y) == 4
> s::y at offset 4
> sizeof(s::y) == 8 (padding, this is what catches people out)
> s::z at offset 4+8 = 12
> sizeof(struct s) == 16 (and here too)
>
> Adding packed attribute on outer struct reduces sizeof(struct s) to 12
> on metag:
> alignof(s::y) == 4
> s::y at offset 2 (packed)
> sizeof(s::y) == 8 (still padded)

In my memory, when packed(2), it breaks the C standard (although I am
not quit sure).

And I guess, all C programmers will assume it will be 6 when within
pack(2) or pack(1).

> s::z at offset 2+8 = 10
> sizeof(struct s) == 12 (packed)
>
> Also reduced to 12 if only inner struct is marked packed:
> alignof(s::y) == 2
> s::y at offset 2
> sizeof(s::y) == 6 (packed)
> s::z at offset 2+6 = 8
> sizeof(struct s) == 12 (still padded)
>
> Adding packed attribute on both outer and inner struct reduces
> sizeof(struct s) to 10 to match x86.
>
> Unfortunately it's years too late to change this ABI, so we're stuck
> with it.
>

Unfortunately too, most using cases are related with API (the related
structure definition must be the same in binary data).

I am sure there are still another ways to bypass this issue, but that
will make the code looks very strange (especially they are API).

:-(

>>
>>
>> Making cross-compiler:
>>
>> git clone git://github.com/img-meta/metag-buildroot.git metag-core-2013.11
>> make meta2_defconfig.
>
> That loads the meta2_defconfig. Did you rebuild it? If in doubt you can
> always "rm -fr output" and re-make.
>

Yeah, I had done, and no doubt to me.


Thanks.
--
Chen Gang

Open, share and attitude like air, water and life which God blessed

2014-01-10 15:57:23

by Chen Gang

[permalink] [raw]
Subject: Re: [Suggest] arch: metag: compiler: Are they compiler's issues?

On 01/08/2014 11:01 PM, Chen Gang wrote:
> On 01/06/2014 06:31 PM, James Hogan wrote:
>> I suspect this is due to bad assumptions in the code. The metag ABI is
>> unusual in padding the size of structs to a 32bit boundary even if all
>> members are <32bit. This is actually permitted by the C standard but
>> it's a bit of a pain. e.g.
>>
>> struct s {
>> short x
>> struct {
>> short x[3];
>> } y;
>> short z;
>> };
>>
>> on x86
>> alignof(s::y) == 2
>> s::y at offset 2
>> sizeof(s::y) == 6
>> s::z at offset 6+2 = 8
>> sizeof(struct s) == 10
>>
>> but on metag
>> alignof(s::y) == 4
>> s::y at offset 4
>> sizeof(s::y) == 8 (padding, this is what catches people out)
>> s::z at offset 4+8 = 12
>> sizeof(struct s) == 16 (and here too)
>>
>> Adding packed attribute on outer struct reduces sizeof(struct s) to 12
>> on metag:
>> alignof(s::y) == 4
>> s::y at offset 2 (packed)
>> sizeof(s::y) == 8 (still padded)
>
> In my memory, when packed(2), it breaks the C standard (although I am
> not quit sure).
>
> And I guess, all C programmers will assume it will be 6 when within
> pack(2) or pack(1).
>
>> s::z at offset 2+8 = 10
>> sizeof(struct s) == 12 (packed)
>>
>> Also reduced to 12 if only inner struct is marked packed:
>> alignof(s::y) == 2
>> s::y at offset 2
>> sizeof(s::y) == 6 (packed)
>> s::z at offset 2+6 = 8
>> sizeof(struct s) == 12 (still padded)
>>
>> Adding packed attribute on both outer and inner struct reduces
>> sizeof(struct s) to 10 to match x86.
>>
>> Unfortunately it's years too late to change this ABI, so we're stuck
>> with it.
>>
>
> Unfortunately too, most using cases are related with API (the related
> structure definition must be the same in binary data).
>
> I am sure there are still another ways to bypass this issue, but that
> will make the code looks very strange (especially they are API).
>
> :-(
>

I guess most C programmers will use this way to describe protocol/data
format, and keep compatible for it (since it is API).

So even if it really does not break C standard, I still recommend our
compiler to improve itself to support this features.


Thanks.
--
Chen Gang

Open, share and attitude like air, water and life which God blessed

2014-01-10 16:02:49

by James Hogan

[permalink] [raw]
Subject: Re: [Suggest] arch: metag: compiler: Are they compiler's issues?

On 10/01/14 15:57, Chen Gang wrote:
> On 01/08/2014 11:01 PM, Chen Gang wrote:
>> On 01/06/2014 06:31 PM, James Hogan wrote:
>>> I suspect this is due to bad assumptions in the code. The metag ABI is
>>> unusual in padding the size of structs to a 32bit boundary even if all
>>> members are <32bit. This is actually permitted by the C standard but
>>> it's a bit of a pain. e.g.
>>>
>>> struct s {
>>> short x
>>> struct {
>>> short x[3];
>>> } y;
>>> short z;
>>> };
>>>
>>> on x86
>>> alignof(s::y) == 2
>>> s::y at offset 2
>>> sizeof(s::y) == 6
>>> s::z at offset 6+2 = 8
>>> sizeof(struct s) == 10
>>>
>>> but on metag
>>> alignof(s::y) == 4
>>> s::y at offset 4
>>> sizeof(s::y) == 8 (padding, this is what catches people out)
>>> s::z at offset 4+8 = 12
>>> sizeof(struct s) == 16 (and here too)
>>>
>>> Adding packed attribute on outer struct reduces sizeof(struct s) to 12
>>> on metag:
>>> alignof(s::y) == 4
>>> s::y at offset 2 (packed)
>>> sizeof(s::y) == 8 (still padded)
>>
>> In my memory, when packed(2), it breaks the C standard (although I am
>> not quit sure).
>>
>> And I guess, all C programmers will assume it will be 6 when within
>> pack(2) or pack(1).
>>
>>> s::z at offset 2+8 = 10
>>> sizeof(struct s) == 12 (packed)
>>>
>>> Also reduced to 12 if only inner struct is marked packed:
>>> alignof(s::y) == 2
>>> s::y at offset 2
>>> sizeof(s::y) == 6 (packed)
>>> s::z at offset 2+6 = 8
>>> sizeof(struct s) == 12 (still padded)
>>>
>>> Adding packed attribute on both outer and inner struct reduces
>>> sizeof(struct s) to 10 to match x86.
>>>
>>> Unfortunately it's years too late to change this ABI, so we're stuck
>>> with it.
>>>
>>
>> Unfortunately too, most using cases are related with API (the related
>> structure definition must be the same in binary data).
>>
>> I am sure there are still another ways to bypass this issue, but that
>> will make the code looks very strange (especially they are API).
>>
>> :-(
>>
>
> I guess most C programmers will use this way to describe protocol/data
> format, and keep compatible for it (since it is API).
>
> So even if it really does not break C standard, I still recommend our
> compiler to improve itself to support this features.

The compiler cannot change this without breaking the ABI.

If the structure describes a set-in-stone data layout (which it sounds
like it does since it asserts the size of it) then the correct fix is to
pack the structures in such a way as to guarantee the correct offsets
and sizes on all compliant compilers. Otherwise if it's just an internal
programming API it shouldn't be using compile time asserts to enforce
things that vary between ABIs.

Cheers
James

2014-01-10 16:20:17

by Chen Gang

[permalink] [raw]
Subject: Re: [Suggest] arch: metag: compiler: Are they compiler's issues?

On 01/11/2014 12:02 AM, James Hogan wrote:
> On 10/01/14 15:57, Chen Gang wrote:
>> On 01/08/2014 11:01 PM, Chen Gang wrote:
>>> On 01/06/2014 06:31 PM, James Hogan wrote:
>>>> I suspect this is due to bad assumptions in the code. The metag ABI is
>>>> unusual in padding the size of structs to a 32bit boundary even if all
>>>> members are <32bit. This is actually permitted by the C standard but
>>>> it's a bit of a pain. e.g.
>>>>
>>>> struct s {
>>>> short x
>>>> struct {
>>>> short x[3];
>>>> } y;
>>>> short z;
>>>> };
>>>>
>>>> on x86
>>>> alignof(s::y) == 2
>>>> s::y at offset 2
>>>> sizeof(s::y) == 6
>>>> s::z at offset 6+2 = 8
>>>> sizeof(struct s) == 10
>>>>
>>>> but on metag
>>>> alignof(s::y) == 4
>>>> s::y at offset 4
>>>> sizeof(s::y) == 8 (padding, this is what catches people out)
>>>> s::z at offset 4+8 = 12
>>>> sizeof(struct s) == 16 (and here too)
>>>>
>>>> Adding packed attribute on outer struct reduces sizeof(struct s) to 12
>>>> on metag:
>>>> alignof(s::y) == 4
>>>> s::y at offset 2 (packed)
>>>> sizeof(s::y) == 8 (still padded)
>>>
>>> In my memory, when packed(2), it breaks the C standard (although I am
>>> not quit sure).
>>>
>>> And I guess, all C programmers will assume it will be 6 when within
>>> pack(2) or pack(1).
>>>
>>>> s::z at offset 2+8 = 10
>>>> sizeof(struct s) == 12 (packed)
>>>>
>>>> Also reduced to 12 if only inner struct is marked packed:
>>>> alignof(s::y) == 2
>>>> s::y at offset 2
>>>> sizeof(s::y) == 6 (packed)
>>>> s::z at offset 2+6 = 8
>>>> sizeof(struct s) == 12 (still padded)
>>>>
>>>> Adding packed attribute on both outer and inner struct reduces
>>>> sizeof(struct s) to 10 to match x86.
>>>>
>>>> Unfortunately it's years too late to change this ABI, so we're stuck
>>>> with it.
>>>>
>>>
>>> Unfortunately too, most using cases are related with API (the related
>>> structure definition must be the same in binary data).
>>>
>>> I am sure there are still another ways to bypass this issue, but that
>>> will make the code looks very strange (especially they are API).
>>>
>>> :-(
>>>
>>
>> I guess most C programmers will use this way to describe protocol/data
>> format, and keep compatible for it (since it is API).
>>
>> So even if it really does not break C standard, I still recommend our
>> compiler to improve itself to support this features.
>
> The compiler cannot change this without breaking the ABI.
>
> If the structure describes a set-in-stone data layout (which it sounds
> like it does since it asserts the size of it) then the correct fix is to
> pack the structures in such a way as to guarantee the correct offsets
> and sizes on all compliant compilers. Otherwise if it's just an internal
> programming API it shouldn't be using compile time asserts to enforce
> things that vary between ABIs.
>

OK, thanks, I guess your meaning is:

struct s {
short x;
struct {
short x[3];
} y __attribute__ ((packed));
short z;
} __attribute__ ((packed));

That will satisfy all of compilers (include metag), is it correct?


Thanks.
--
Chen Gang

Open, share and attitude like air, water and life which God blessed

2014-01-10 16:31:30

by James Hogan

[permalink] [raw]
Subject: Re: [Suggest] arch: metag: compiler: Are they compiler's issues?

On 10/01/14 16:20, Chen Gang wrote:
> On 01/11/2014 12:02 AM, James Hogan wrote:
>> On 10/01/14 15:57, Chen Gang wrote:
>>> On 01/08/2014 11:01 PM, Chen Gang wrote:
>>>> On 01/06/2014 06:31 PM, James Hogan wrote:
>>>>> I suspect this is due to bad assumptions in the code. The metag ABI is
>>>>> unusual in padding the size of structs to a 32bit boundary even if all
>>>>> members are <32bit. This is actually permitted by the C standard but
>>>>> it's a bit of a pain. e.g.
>>>>>
>>>>> struct s {
>>>>> short x
>>>>> struct {
>>>>> short x[3];
>>>>> } y;
>>>>> short z;
>>>>> };
>>>>>
>>>>> on x86
>>>>> alignof(s::y) == 2
>>>>> s::y at offset 2
>>>>> sizeof(s::y) == 6
>>>>> s::z at offset 6+2 = 8
>>>>> sizeof(struct s) == 10
>>>>>
>>>>> but on metag
>>>>> alignof(s::y) == 4
>>>>> s::y at offset 4
>>>>> sizeof(s::y) == 8 (padding, this is what catches people out)
>>>>> s::z at offset 4+8 = 12
>>>>> sizeof(struct s) == 16 (and here too)
>>>>>
>>>>> Adding packed attribute on outer struct reduces sizeof(struct s) to 12
>>>>> on metag:
>>>>> alignof(s::y) == 4
>>>>> s::y at offset 2 (packed)
>>>>> sizeof(s::y) == 8 (still padded)
>>>>
>>>> In my memory, when packed(2), it breaks the C standard (although I am
>>>> not quit sure).
>>>>
>>>> And I guess, all C programmers will assume it will be 6 when within
>>>> pack(2) or pack(1).
>>>>
>>>>> s::z at offset 2+8 = 10
>>>>> sizeof(struct s) == 12 (packed)
>>>>>
>>>>> Also reduced to 12 if only inner struct is marked packed:
>>>>> alignof(s::y) == 2
>>>>> s::y at offset 2
>>>>> sizeof(s::y) == 6 (packed)
>>>>> s::z at offset 2+6 = 8
>>>>> sizeof(struct s) == 12 (still padded)
>>>>>
>>>>> Adding packed attribute on both outer and inner struct reduces
>>>>> sizeof(struct s) to 10 to match x86.
>>>>>
>>>>> Unfortunately it's years too late to change this ABI, so we're stuck
>>>>> with it.
>>>>>
>>>>
>>>> Unfortunately too, most using cases are related with API (the related
>>>> structure definition must be the same in binary data).
>>>>
>>>> I am sure there are still another ways to bypass this issue, but that
>>>> will make the code looks very strange (especially they are API).
>>>>
>>>> :-(
>>>>
>>>
>>> I guess most C programmers will use this way to describe protocol/data
>>> format, and keep compatible for it (since it is API).
>>>
>>> So even if it really does not break C standard, I still recommend our
>>> compiler to improve itself to support this features.
>>
>> The compiler cannot change this without breaking the ABI.
>>
>> If the structure describes a set-in-stone data layout (which it sounds
>> like it does since it asserts the size of it) then the correct fix is to
>> pack the structures in such a way as to guarantee the correct offsets
>> and sizes on all compliant compilers. Otherwise if it's just an internal
>> programming API it shouldn't be using compile time asserts to enforce
>> things that vary between ABIs.
>>
>
> OK, thanks, I guess your meaning is:
>
> struct s {
> short x;
> struct {
> short x[3];
> } y __attribute__ ((packed));
> short z;
> } __attribute__ ((packed));
>
> That will satisfy all of compilers (include metag), is it correct?

Yes, that's what I mean (although probably best to use the __packed
macro rather than __attribute__ ((packed)) ).

Cheers
James

2014-01-10 16:38:55

by Chen Gang

[permalink] [raw]
Subject: Re: [Suggest] arch: metag: compiler: Are they compiler's issues?

On 01/11/2014 12:30 AM, James Hogan wrote:
> On 10/01/14 16:20, Chen Gang wrote:
>> On 01/11/2014 12:02 AM, James Hogan wrote:
>>> On 10/01/14 15:57, Chen Gang wrote:
>>>> On 01/08/2014 11:01 PM, Chen Gang wrote:
>>>>> On 01/06/2014 06:31 PM, James Hogan wrote:
>>>>>> I suspect this is due to bad assumptions in the code. The metag ABI is
>>>>>> unusual in padding the size of structs to a 32bit boundary even if all
>>>>>> members are <32bit. This is actually permitted by the C standard but
>>>>>> it's a bit of a pain. e.g.
>>>>>>
>>>>>> struct s {
>>>>>> short x
>>>>>> struct {
>>>>>> short x[3];
>>>>>> } y;
>>>>>> short z;
>>>>>> };
>>>>>>
>>>>>> on x86
>>>>>> alignof(s::y) == 2
>>>>>> s::y at offset 2
>>>>>> sizeof(s::y) == 6
>>>>>> s::z at offset 6+2 = 8
>>>>>> sizeof(struct s) == 10
>>>>>>
>>>>>> but on metag
>>>>>> alignof(s::y) == 4
>>>>>> s::y at offset 4
>>>>>> sizeof(s::y) == 8 (padding, this is what catches people out)
>>>>>> s::z at offset 4+8 = 12
>>>>>> sizeof(struct s) == 16 (and here too)
>>>>>>
>>>>>> Adding packed attribute on outer struct reduces sizeof(struct s) to 12
>>>>>> on metag:
>>>>>> alignof(s::y) == 4
>>>>>> s::y at offset 2 (packed)
>>>>>> sizeof(s::y) == 8 (still padded)
>>>>>
>>>>> In my memory, when packed(2), it breaks the C standard (although I am
>>>>> not quit sure).
>>>>>
>>>>> And I guess, all C programmers will assume it will be 6 when within
>>>>> pack(2) or pack(1).
>>>>>
>>>>>> s::z at offset 2+8 = 10
>>>>>> sizeof(struct s) == 12 (packed)
>>>>>>
>>>>>> Also reduced to 12 if only inner struct is marked packed:
>>>>>> alignof(s::y) == 2
>>>>>> s::y at offset 2
>>>>>> sizeof(s::y) == 6 (packed)
>>>>>> s::z at offset 2+6 = 8
>>>>>> sizeof(struct s) == 12 (still padded)
>>>>>>
>>>>>> Adding packed attribute on both outer and inner struct reduces
>>>>>> sizeof(struct s) to 10 to match x86.
>>>>>>
>>>>>> Unfortunately it's years too late to change this ABI, so we're stuck
>>>>>> with it.
>>>>>>
>>>>>
>>>>> Unfortunately too, most using cases are related with API (the related
>>>>> structure definition must be the same in binary data).
>>>>>
>>>>> I am sure there are still another ways to bypass this issue, but that
>>>>> will make the code looks very strange (especially they are API).
>>>>>
>>>>> :-(
>>>>>
>>>>
>>>> I guess most C programmers will use this way to describe protocol/data
>>>> format, and keep compatible for it (since it is API).
>>>>
>>>> So even if it really does not break C standard, I still recommend our
>>>> compiler to improve itself to support this features.
>>>
>>> The compiler cannot change this without breaking the ABI.
>>>
>>> If the structure describes a set-in-stone data layout (which it sounds
>>> like it does since it asserts the size of it) then the correct fix is to
>>> pack the structures in such a way as to guarantee the correct offsets
>>> and sizes on all compliant compilers. Otherwise if it's just an internal
>>> programming API it shouldn't be using compile time asserts to enforce
>>> things that vary between ABIs.
>>>
>>
>> OK, thanks, I guess your meaning is:
>>
>> struct s {
>> short x;
>> struct {
>> short x[3];
>> } y __attribute__ ((packed));
>> short z;
>> } __attribute__ ((packed));
>>
>> That will satisfy all of compilers (include metag), is it correct?
>
> Yes, that's what I mean (although probably best to use the __packed
> macro rather than __attribute__ ((packed)) ).
>

OK, thanks, and excuse me, during these days, I have no quite enough
time for upstream kernel.

So, I plan that I will/should send related patches for it within next
week end (2014-01-19), if it is too long to bear it, please help send
related patches for it, thanks.


Thanks.
--
Chen Gang

Open, share and attitude like air, water and life which God blessed