Re: [kvm-unit-tests PATCH 1/5] s390x: Add a linker script to assembly snippets

[Janosch Frank <frankja@linux.ibm.com>]

On 11/23/22 13:45, Claudio Imbrenda wrote:
> On Wed, 23 Nov 2022 08:46:52 +0000
> Janosch Frank <frankja@linux.ibm.com> wrote:
> 
>> A linker script has a few benefits:
>> - Random data doesn't end up in the binary breaking tests
>> - We can easily define a lowcore and load the snippet from 0x0 instead
>> of 0x4000 which makes asm snippets behave like c snippets
>> - We can easily define an invalid PGM new PSW to ensure an exit on a
>> guest PGM
>>
>> Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
> 
> looks good in general, but I have a few questions
> 
>> ---
>>   lib/s390x/snippet.h         |  3 +--
>>   s390x/Makefile              |  5 +++--
>>   s390x/mvpg-sie.c            |  2 +-
>>   s390x/pv-diags.c            |  6 +++---
>>   s390x/snippets/asm/flat.lds | 43 +++++++++++++++++++++++++++++++++++++
>>   5 files changed, 51 insertions(+), 8 deletions(-)
>>   create mode 100644 s390x/snippets/asm/flat.lds
>>
>> diff --git a/lib/s390x/snippet.h b/lib/s390x/snippet.h
>> index b17b2a4c..57045994 100644
>> --- a/lib/s390x/snippet.h
>> +++ b/lib/s390x/snippet.h
>> @@ -32,8 +32,7 @@
>>   
>>   #define SNIPPET_PV_TWEAK0	0x42UL
>>   #define SNIPPET_PV_TWEAK1	0UL
>> -#define SNIPPET_OFF_C		0
>> -#define SNIPPET_OFF_ASM		0x4000
>> +#define SNIPPET_UNPACK_OFF	0
>>   
>>   
>>   /*
>> diff --git a/s390x/Makefile b/s390x/Makefile
>> index bf1504f9..bb0f9eb8 100644
>> --- a/s390x/Makefile
>> +++ b/s390x/Makefile
>> @@ -135,7 +135,8 @@ $(SNIPPET_DIR)/c/%.o: $(SNIPPET_DIR)/c/%.c $(asm-offsets)
>>   	$(CC) $(CFLAGS) -c -nostdlib -o $@ $<
>>   
>>   $(SNIPPET_DIR)/asm/%.gbin: $(SNIPPET_DIR)/asm/%.o
>> -	$(OBJCOPY) -O binary -j ".rodata" -j ".text" -j ".data" -j ".bss" --set-section-flags .bss=alloc,load,contents $(patsubst %.gbin,%.o,$@) $@
>> +	$(CC) $(LDFLAGS) -o $@ -T $(SRCDIR)/s390x/snippets/asm/flat.lds $(patsubst %.gbin,%.o,$@)
> 
> I think you can simply use $< instead of the patsubst expression

Right, I'll fix that momentarily.

[...]

>> +	.text : {
>> +		*(.text)
>> +		*(.text.*)
>> +	}
>> +	. = ALIGN(64K);
> 
> any reason to align to 64k? (instead of e.g. 4k)

Well, I've copied that from s390x/flat.lds...
I'll have a look at the required alignments when I find time.

> 
>> +	etext = .;
>> +	. = ALIGN(16);
> 
> do you need the ALIGN? I would think we are already aligned here
> 
>> +	.data : {
>> +		*(.data)
>> +		*(.data.rel*)
>> +	}
>> +	. = ALIGN(16);
>> +	.rodata : { *(.rodata) *(.rodata.*) }
>> +	. = ALIGN(16);
>> +	__bss_start = .;
>> +	.bss : { *(.bss) }
>> +	__bss_end = .;
>> +	. = ALIGN(64K);
> 
> same question as above regarding 64k
> 
>> +}
>