public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Claudio Imbrenda <imbrenda@linux.ibm.com>
To: Janosch Frank <frankja@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: Thu, 12 Jan 2023 18:13:53 +0100	[thread overview]
Message-ID: <20230112181353.0ed31250@p-imbrenda> (raw)
In-Reply-To: <20230112154548.163021-8-frankja@linux.ibm.com>

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)

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:[~2023-01-12 17:54 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 [this message]
2023-01-13 15:50     ` Janosch Frank

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=20230112181353.0ed31250@p-imbrenda \
    --to=imbrenda@linux.ibm.com \
    --cc=david@redhat.com \
    --cc=frankja@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