linux-s390.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Janosch Frank <frankja@linux.ibm.com>
To: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: kvm@vger.kernel.org, linux-s390@vger.kernel.org,
	thuth@redhat.com, david@redhat.com, nsg@linux.ibm.com,
	nrb@linux.ibm.com
Subject: Re: [kvm-unit-tests PATCH v2 7/7] lib: s390x: Handle debug prints for SIE exceptions correctly
Date: Fri, 13 Jan 2023 16:50:51 +0100	[thread overview]
Message-ID: <54e21150-714c-4a06-59ba-a599bae76bbf@linux.ibm.com> (raw)
In-Reply-To: <20230112181353.0ed31250@p-imbrenda>

On 1/12/23 18:13, Claudio Imbrenda wrote:
> On Thu, 12 Jan 2023 15:45:48 +0000
> Janosch Frank <frankja@linux.ibm.com> wrote:
> 
>> When we leave SIE due to an exception, we'll still have guest values
>> in registers 0 - 13 and that's not clearly portraied in our debug
>> prints. So let's fix that.
>>
>> Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
>> ---
>>   lib/s390x/interrupt.c | 46 ++++++++++++++++++++++++++++++++++++++-----
>>   lib/s390x/sie.h       |  2 ++
>>   s390x/cpu.S           |  6 ++++--
>>   3 files changed, 47 insertions(+), 7 deletions(-)
>>
>> diff --git a/lib/s390x/interrupt.c b/lib/s390x/interrupt.c
>> index dadb7415..ff47c2c2 100644
>> --- a/lib/s390x/interrupt.c
>> +++ b/lib/s390x/interrupt.c
>> @@ -9,6 +9,7 @@
>>    */
>>   #include <libcflat.h>
>>   #include <asm/barrier.h>
>> +#include <asm/asm-offsets.h>
>>   #include <sclp.h>
>>   #include <interrupt.h>
>>   #include <sie.h>
>> @@ -188,9 +189,12 @@ static void print_storage_exception_information(void)
>>   	}
>>   }
>>   
>> -static void print_int_regs(struct stack_frame_int *stack)
>> +static void print_int_regs(struct stack_frame_int *stack, bool sie)
>>   {
>> +	struct kvm_s390_sie_block *sblk;
>> +
>>   	printf("\n");
>> +	printf("%s\n", sie ? "Guest registers:" : "Host registers:");
>>   	printf("GPRS:\n");
>>   	printf("%016lx %016lx %016lx %016lx\n",
>>   	       stack->grs1[0], stack->grs1[1], stack->grs0[0], stack->grs0[1]);
>> @@ -198,24 +202,56 @@ static void print_int_regs(struct stack_frame_int *stack)
>>   	       stack->grs0[2], stack->grs0[3], stack->grs0[4], stack->grs0[5]);
>>   	printf("%016lx %016lx %016lx %016lx\n",
>>   	       stack->grs0[6], stack->grs0[7], stack->grs0[8], stack->grs0[9]);
>> -	printf("%016lx %016lx %016lx %016lx\n",
>> -	       stack->grs0[10], stack->grs0[11], stack->grs0[12], stack->grs0[13]);
>> +
>> +	if (sie) {
>> +		sblk = (struct kvm_s390_sie_block *)stack->grs0[12];
>> +		printf("%016lx %016lx %016lx %016lx\n",
>> +		       stack->grs0[10], stack->grs0[11], sblk->gg14, sblk->gg15);
>> +	} else {
>> +		printf("%016lx %016lx %016lx %016lx\n",
>> +		       stack->grs0[10], stack->grs0[11], stack->grs0[12], stack->grs0[13]);
>> +	}
>> +
>>   	printf("\n");
>>   }
>>   
>>   static void print_pgm_info(struct stack_frame_int *stack)
>>   
>>   {
>> -	bool in_sie;
>> +	bool in_sie, in_sie_gregs;
>> +	struct vm_save_area *vregs;
>>   
>>   	in_sie = (lowcore.pgm_old_psw.addr >= (uintptr_t)sie_entry &&
>>   		  lowcore.pgm_old_psw.addr <= (uintptr_t)sie_exit);
>> +	in_sie_gregs = (lowcore.pgm_old_psw.addr >= (uintptr_t)sie_entry_gregs &&
>> +			lowcore.pgm_old_psw.addr <= (uintptr_t)sie_exit_gregs);
> 
> can you explain why <= instead of < ? (I think I know why, but a
> comment would not hurt)

This might be wrong now that the exit label points behind a lmg which 
can only cause the operation or the addressing exception.

The operation exception can only happen on a LMY without the long 
displacement HW support and the addressing exception should only happen 
in circumstances where any processed instruction will likely result in a 
new exception being thrown.

> 
> with that fixed:
> 
> Reviewed-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
> 
>>   
>>   	printf("\n");
>>   	printf("Unexpected program interrupt %s: %#x on cpu %d at %#lx, ilen %d\n",
>>   	       in_sie ? "in SIE" : "",
>>   	       lowcore.pgm_int_code, stap(), lowcore.pgm_old_psw.addr, lowcore.pgm_int_id);
>> -	print_int_regs(stack);
>> +
>> +	/*
>> +	 * If we fall out of SIE before loading the host registers,
>> +	 * then we need to do it here so we print the host registers
>> +	 * and not the guest registers.
>> +	 *
>> +	 * Back tracing is actually not a problem since SIE restores gr15.
>> +	 */
>> +	if (in_sie_gregs) {
>> +		print_int_regs(stack, true);
>> +		vregs = *((struct vm_save_area **)(stack->grs0[13] + __SF_SIE_SAVEAREA));
>> +
>> +		/*
>> +		 * The grs are not linear on the interrupt stack frame.
>> +		 * We copy 0 and 1 here and 2 - 15 with the memcopy below.
>> +		 */
>> +		stack->grs1[0] = vregs->host.grs[0];
>> +		stack->grs1[1] = vregs->host.grs[1];
>> +		/*  2 - 15 */
>> +		memcpy(stack->grs0, &vregs->host.grs[2], sizeof(stack->grs0) - 8);
>> +	}
>> +	print_int_regs(stack, false);
>>   	dump_stack();
>>   
>>   	/* Dump stack doesn't end with a \n so we add it here instead */
>> diff --git a/lib/s390x/sie.h b/lib/s390x/sie.h
>> index a27a8401..147cb0f2 100644
>> --- a/lib/s390x/sie.h
>> +++ b/lib/s390x/sie.h
>> @@ -273,6 +273,8 @@ struct vm {
>>   
>>   extern void sie_entry(void);
>>   extern void sie_exit(void);
>> +extern void sie_entry_gregs(void);
>> +extern void sie_exit_gregs(void);
>>   extern void sie64a(struct kvm_s390_sie_block *sblk, struct vm_save_area *save_area);
>>   void sie(struct vm *vm);
>>   void sie_expect_validity(struct vm *vm);
>> diff --git a/s390x/cpu.S b/s390x/cpu.S
>> index 45bd551a..9155b044 100644
>> --- a/s390x/cpu.S
>> +++ b/s390x/cpu.S
>> @@ -82,7 +82,8 @@ sie64a:
>>   	# Store scb and save_area pointer into stack frame
>>   	stg	%r2,__SF_SIE_CONTROL(%r15)	# save control block pointer
>>   	stg	%r3,__SF_SIE_SAVEAREA(%r15)	# save guest register save area
>> -
>> +.globl sie_entry_gregs
>> +sie_entry_gregs:
>>   	# Load guest's gprs, fprs and fpc
>>   	.irp i, 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15
>>   	ld	\i, \i * 8 + SIE_SAVEAREA_GUEST_FPRS(%r3)
>> @@ -121,7 +122,8 @@ sie_exit:
>>   	.endr
>>   	lfpc	SIE_SAVEAREA_HOST_FPC(%r14)
>>   	lmg	%r0,%r14,SIE_SAVEAREA_HOST_GRS(%r14)	# restore kernel registers
>> -
>> +.globl sie_exit_gregs
>> +sie_exit_gregs:
>>   	br	%r14
>>   
>>   	.align	8
> 


      reply	other threads:[~2023-01-13 16:00 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-12 15:45 [kvm-unit-tests PATCH v2 0/7] s390x: Snippet fixes Janosch Frank
2023-01-12 15:45 ` [kvm-unit-tests PATCH v2 1/7] s390x: Cleanup flat.lds Janosch Frank
2023-01-12 17:05   ` Claudio Imbrenda
2023-01-12 15:45 ` [kvm-unit-tests PATCH v2 2/7] s390x: snippets: c: " Janosch Frank
2023-01-12 17:05   ` Claudio Imbrenda
2023-01-12 15:45 ` [kvm-unit-tests PATCH v2 3/7] s390x: Add a linker script to assembly snippets Janosch Frank
2023-01-12 17:05   ` Claudio Imbrenda
2023-01-12 15:45 ` [kvm-unit-tests PATCH v2 4/7] s390x: snippets: Fix SET_PSW_NEW_ADDR macro Janosch Frank
2023-01-12 15:45 ` [kvm-unit-tests PATCH v2 5/7] lib: s390x: sie: Set guest memory pointer Janosch Frank
2023-01-12 15:45 ` [kvm-unit-tests PATCH v2 6/7] s390x: Clear first stack frame and end backtrace early Janosch Frank
2023-01-12 15:45 ` [kvm-unit-tests PATCH v2 7/7] lib: s390x: Handle debug prints for SIE exceptions correctly Janosch Frank
2023-01-12 17:13   ` Claudio Imbrenda
2023-01-13 15:50     ` Janosch Frank [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=54e21150-714c-4a06-59ba-a599bae76bbf@linux.ibm.com \
    --to=frankja@linux.ibm.com \
    --cc=david@redhat.com \
    --cc=imbrenda@linux.ibm.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=nrb@linux.ibm.com \
    --cc=nsg@linux.ibm.com \
    --cc=thuth@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).