From: Claudio Imbrenda <imbrenda@linux.ibm.com>
To: Janosch Frank <frankja@linux.ibm.com>
Cc: kvm@vger.kernel.org, linux-s390@vger.kernel.org,
david@redhat.com, thuth@redhat.com, seiden@linux.ibm.com,
nrb@linux.ibm.com
Subject: Re: [kvm-unit-tests PATCH 5/5] lib: s390x: Handle debug prints for SIE exceptions correctly
Date: Wed, 23 Nov 2022 14:01:18 +0100 [thread overview]
Message-ID: <20221123140118.25114940@p-imbrenda> (raw)
In-Reply-To: <20221123084656.19864-6-frankja@linux.ibm.com>
On Wed, 23 Nov 2022 08:46:56 +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.
wouldn't it be cleaner to restore the registers in the interrupt
handler? (I thought we were already doing it)
>
> 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);
>
> 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
next prev parent reply other threads:[~2022-11-23 13:28 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-23 8:46 [kvm-unit-tests PATCH 0/5] s390x: Snippet fixes Janosch Frank
2022-11-23 8:46 ` [kvm-unit-tests PATCH 1/5] s390x: Add a linker script to assembly snippets Janosch Frank
2022-11-23 12:45 ` Claudio Imbrenda
2022-11-24 14:28 ` Janosch Frank
2022-11-24 20:42 ` Janis Schoetterl-Glausch
2022-11-25 10:07 ` Janosch Frank
2022-11-23 8:46 ` [kvm-unit-tests PATCH 2/5] s390x: snippets: Fix SET_PSW_NEW_ADDR macro Janosch Frank
2022-11-23 10:58 ` Nico Boehr
2022-11-23 13:05 ` Claudio Imbrenda
2022-11-23 8:46 ` [kvm-unit-tests PATCH 3/5] lib: s390x: sie: Set guest memory pointer Janosch Frank
2022-11-23 11:01 ` Nico Boehr
2022-11-23 12:46 ` Claudio Imbrenda
2022-11-23 8:46 ` [kvm-unit-tests PATCH 4/5] s390x: Clear first stack frame and end backtrace early Janosch Frank
2022-11-23 11:05 ` Nico Boehr
2022-11-23 12:47 ` Claudio Imbrenda
2022-11-23 8:46 ` [kvm-unit-tests PATCH 5/5] lib: s390x: Handle debug prints for SIE exceptions correctly Janosch Frank
2022-11-23 13:01 ` Claudio Imbrenda [this message]
2022-11-30 14:38 ` Janosch Frank
2022-11-30 14:46 ` Claudio Imbrenda
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=20221123140118.25114940@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=seiden@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