From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: Juan Quintela <quintela@redhat.com>,
patches@linaro.org, qemu-devel@nongnu.org,
Shannon Zhao <shannon.zhaosl@gmail.com>,
qemu-arm@nongnu.org, Shannon Zhao <zhaoshenglong@huawei.com>
Subject: Re: [Qemu-arm] [PATCH for-3.0 v2 5/5] hw/intc/arm_gicv3_common: Move gicd shift bug handling to gicv3_post_load
Date: Mon, 6 Aug 2018 16:39:25 +0100 [thread overview]
Message-ID: <20180806153925.GF2473@work-vm> (raw)
In-Reply-To: <20180806123445.1459-6-peter.maydell@linaro.org>
* Peter Maydell (peter.maydell@linaro.org) wrote:
> The code currently in gicv3_gicd_no_migration_shift_bug_post_load()
> that handles migration from older QEMU versions with a particular
> bug is misplaced. We need to run this after migration in all cases,
> not just the cases where the "arm_gicv3/gicd_no_migration_shift_bug"
> subsection is present, so it must go in a post_load hook for the
> top level VMSD, not for the subsection. Move it.
>
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
> ---
> hw/intc/arm_gicv3_common.c | 77 ++++++++++++++++++--------------------
> 1 file changed, 37 insertions(+), 40 deletions(-)
>
> diff --git a/hw/intc/arm_gicv3_common.c b/hw/intc/arm_gicv3_common.c
> index 8175889f1e7..52480c3b4cf 100644
> --- a/hw/intc/arm_gicv3_common.c
> +++ b/hw/intc/arm_gicv3_common.c
> @@ -29,6 +29,41 @@
> #include "hw/arm/linux-boot-if.h"
> #include "sysemu/kvm.h"
>
> +
> +static void gicv3_gicd_no_migration_shift_bug_post_load(GICv3State *cs)
> +{
> + if (cs->gicd_no_migration_shift_bug) {
> + return;
> + }
> +
> + /* Older versions of QEMU had a bug in the handling of state save/restore
> + * to the KVM GICv3: they got the offset in the bitmap arrays wrong,
> + * so that instead of the data for external interrupts 32 and up
> + * starting at bit position 32 in the bitmap, it started at bit
> + * position 64. If we're receiving data from a QEMU with that bug,
> + * we must move the data down into the right place.
> + */
> + memmove(cs->group, (uint8_t *)cs->group + GIC_INTERNAL / 8,
> + sizeof(cs->group) - GIC_INTERNAL / 8);
> + memmove(cs->grpmod, (uint8_t *)cs->grpmod + GIC_INTERNAL / 8,
> + sizeof(cs->grpmod) - GIC_INTERNAL / 8);
> + memmove(cs->enabled, (uint8_t *)cs->enabled + GIC_INTERNAL / 8,
> + sizeof(cs->enabled) - GIC_INTERNAL / 8);
> + memmove(cs->pending, (uint8_t *)cs->pending + GIC_INTERNAL / 8,
> + sizeof(cs->pending) - GIC_INTERNAL / 8);
> + memmove(cs->active, (uint8_t *)cs->active + GIC_INTERNAL / 8,
> + sizeof(cs->active) - GIC_INTERNAL / 8);
> + memmove(cs->edge_trigger, (uint8_t *)cs->edge_trigger + GIC_INTERNAL / 8,
> + sizeof(cs->edge_trigger) - GIC_INTERNAL / 8);
> +
> + /*
> + * While this new version QEMU doesn't have this kind of bug as we fix it,
> + * so it needs to set the flag to true to indicate that and it's necessary
> + * for next migration to work from this new version QEMU.
> + */
> + cs->gicd_no_migration_shift_bug = true;
> +}
> +
> static int gicv3_pre_save(void *opaque)
> {
> GICv3State *s = (GICv3State *)opaque;
> @@ -46,6 +81,8 @@ static int gicv3_post_load(void *opaque, int version_id)
> GICv3State *s = (GICv3State *)opaque;
> ARMGICv3CommonClass *c = ARM_GICV3_COMMON_GET_CLASS(s);
>
> + gicv3_gicd_no_migration_shift_bug_post_load(s);
> +
> if (c->post_load) {
> c->post_load(s);
> }
> @@ -161,45 +198,6 @@ static int gicv3_pre_load(void *opaque)
> return 0;
> }
>
> -static int gicv3_gicd_no_migration_shift_bug_post_load(void *opaque,
> - int version_id)
> -{
> - GICv3State *cs = opaque;
> -
> - if (cs->gicd_no_migration_shift_bug) {
> - return 0;
> - }
> -
> - /* Older versions of QEMU had a bug in the handling of state save/restore
> - * to the KVM GICv3: they got the offset in the bitmap arrays wrong,
> - * so that instead of the data for external interrupts 32 and up
> - * starting at bit position 32 in the bitmap, it started at bit
> - * position 64. If we're receiving data from a QEMU with that bug,
> - * we must move the data down into the right place.
> - */
> - memmove(cs->group, (uint8_t *)cs->group + GIC_INTERNAL / 8,
> - sizeof(cs->group) - GIC_INTERNAL / 8);
> - memmove(cs->grpmod, (uint8_t *)cs->grpmod + GIC_INTERNAL / 8,
> - sizeof(cs->grpmod) - GIC_INTERNAL / 8);
> - memmove(cs->enabled, (uint8_t *)cs->enabled + GIC_INTERNAL / 8,
> - sizeof(cs->enabled) - GIC_INTERNAL / 8);
> - memmove(cs->pending, (uint8_t *)cs->pending + GIC_INTERNAL / 8,
> - sizeof(cs->pending) - GIC_INTERNAL / 8);
> - memmove(cs->active, (uint8_t *)cs->active + GIC_INTERNAL / 8,
> - sizeof(cs->active) - GIC_INTERNAL / 8);
> - memmove(cs->edge_trigger, (uint8_t *)cs->edge_trigger + GIC_INTERNAL / 8,
> - sizeof(cs->edge_trigger) - GIC_INTERNAL / 8);
> -
> - /*
> - * While this new version QEMU doesn't have this kind of bug as we fix it,
> - * so it needs to set the flag to true to indicate that and it's necessary
> - * for next migration to work from this new version QEMU.
> - */
> - cs->gicd_no_migration_shift_bug = true;
> -
> - return 0;
> -}
> -
> static bool needed_always(void *opaque)
> {
> return true;
> @@ -210,7 +208,6 @@ const VMStateDescription vmstate_gicv3_gicd_no_migration_shift_bug = {
> .version_id = 1,
> .minimum_version_id = 1,
> .needed = needed_always,
> - .post_load = gicv3_gicd_no_migration_shift_bug_post_load,
> .fields = (VMStateField[]) {
> VMSTATE_BOOL(gicd_no_migration_shift_bug, GICv3State),
> VMSTATE_END_OF_LIST()
> --
> 2.17.1
>
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
WARNING: multiple messages have this Message-ID (diff)
From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: qemu-arm@nongnu.org, qemu-devel@nongnu.org, patches@linaro.org,
Juan Quintela <quintela@redhat.com>,
Shannon Zhao <zhaoshenglong@huawei.com>,
Shannon Zhao <shannon.zhaosl@gmail.com>
Subject: Re: [Qemu-devel] [PATCH for-3.0 v2 5/5] hw/intc/arm_gicv3_common: Move gicd shift bug handling to gicv3_post_load
Date: Mon, 6 Aug 2018 16:39:25 +0100 [thread overview]
Message-ID: <20180806153925.GF2473@work-vm> (raw)
In-Reply-To: <20180806123445.1459-6-peter.maydell@linaro.org>
* Peter Maydell (peter.maydell@linaro.org) wrote:
> The code currently in gicv3_gicd_no_migration_shift_bug_post_load()
> that handles migration from older QEMU versions with a particular
> bug is misplaced. We need to run this after migration in all cases,
> not just the cases where the "arm_gicv3/gicd_no_migration_shift_bug"
> subsection is present, so it must go in a post_load hook for the
> top level VMSD, not for the subsection. Move it.
>
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
> ---
> hw/intc/arm_gicv3_common.c | 77 ++++++++++++++++++--------------------
> 1 file changed, 37 insertions(+), 40 deletions(-)
>
> diff --git a/hw/intc/arm_gicv3_common.c b/hw/intc/arm_gicv3_common.c
> index 8175889f1e7..52480c3b4cf 100644
> --- a/hw/intc/arm_gicv3_common.c
> +++ b/hw/intc/arm_gicv3_common.c
> @@ -29,6 +29,41 @@
> #include "hw/arm/linux-boot-if.h"
> #include "sysemu/kvm.h"
>
> +
> +static void gicv3_gicd_no_migration_shift_bug_post_load(GICv3State *cs)
> +{
> + if (cs->gicd_no_migration_shift_bug) {
> + return;
> + }
> +
> + /* Older versions of QEMU had a bug in the handling of state save/restore
> + * to the KVM GICv3: they got the offset in the bitmap arrays wrong,
> + * so that instead of the data for external interrupts 32 and up
> + * starting at bit position 32 in the bitmap, it started at bit
> + * position 64. If we're receiving data from a QEMU with that bug,
> + * we must move the data down into the right place.
> + */
> + memmove(cs->group, (uint8_t *)cs->group + GIC_INTERNAL / 8,
> + sizeof(cs->group) - GIC_INTERNAL / 8);
> + memmove(cs->grpmod, (uint8_t *)cs->grpmod + GIC_INTERNAL / 8,
> + sizeof(cs->grpmod) - GIC_INTERNAL / 8);
> + memmove(cs->enabled, (uint8_t *)cs->enabled + GIC_INTERNAL / 8,
> + sizeof(cs->enabled) - GIC_INTERNAL / 8);
> + memmove(cs->pending, (uint8_t *)cs->pending + GIC_INTERNAL / 8,
> + sizeof(cs->pending) - GIC_INTERNAL / 8);
> + memmove(cs->active, (uint8_t *)cs->active + GIC_INTERNAL / 8,
> + sizeof(cs->active) - GIC_INTERNAL / 8);
> + memmove(cs->edge_trigger, (uint8_t *)cs->edge_trigger + GIC_INTERNAL / 8,
> + sizeof(cs->edge_trigger) - GIC_INTERNAL / 8);
> +
> + /*
> + * While this new version QEMU doesn't have this kind of bug as we fix it,
> + * so it needs to set the flag to true to indicate that and it's necessary
> + * for next migration to work from this new version QEMU.
> + */
> + cs->gicd_no_migration_shift_bug = true;
> +}
> +
> static int gicv3_pre_save(void *opaque)
> {
> GICv3State *s = (GICv3State *)opaque;
> @@ -46,6 +81,8 @@ static int gicv3_post_load(void *opaque, int version_id)
> GICv3State *s = (GICv3State *)opaque;
> ARMGICv3CommonClass *c = ARM_GICV3_COMMON_GET_CLASS(s);
>
> + gicv3_gicd_no_migration_shift_bug_post_load(s);
> +
> if (c->post_load) {
> c->post_load(s);
> }
> @@ -161,45 +198,6 @@ static int gicv3_pre_load(void *opaque)
> return 0;
> }
>
> -static int gicv3_gicd_no_migration_shift_bug_post_load(void *opaque,
> - int version_id)
> -{
> - GICv3State *cs = opaque;
> -
> - if (cs->gicd_no_migration_shift_bug) {
> - return 0;
> - }
> -
> - /* Older versions of QEMU had a bug in the handling of state save/restore
> - * to the KVM GICv3: they got the offset in the bitmap arrays wrong,
> - * so that instead of the data for external interrupts 32 and up
> - * starting at bit position 32 in the bitmap, it started at bit
> - * position 64. If we're receiving data from a QEMU with that bug,
> - * we must move the data down into the right place.
> - */
> - memmove(cs->group, (uint8_t *)cs->group + GIC_INTERNAL / 8,
> - sizeof(cs->group) - GIC_INTERNAL / 8);
> - memmove(cs->grpmod, (uint8_t *)cs->grpmod + GIC_INTERNAL / 8,
> - sizeof(cs->grpmod) - GIC_INTERNAL / 8);
> - memmove(cs->enabled, (uint8_t *)cs->enabled + GIC_INTERNAL / 8,
> - sizeof(cs->enabled) - GIC_INTERNAL / 8);
> - memmove(cs->pending, (uint8_t *)cs->pending + GIC_INTERNAL / 8,
> - sizeof(cs->pending) - GIC_INTERNAL / 8);
> - memmove(cs->active, (uint8_t *)cs->active + GIC_INTERNAL / 8,
> - sizeof(cs->active) - GIC_INTERNAL / 8);
> - memmove(cs->edge_trigger, (uint8_t *)cs->edge_trigger + GIC_INTERNAL / 8,
> - sizeof(cs->edge_trigger) - GIC_INTERNAL / 8);
> -
> - /*
> - * While this new version QEMU doesn't have this kind of bug as we fix it,
> - * so it needs to set the flag to true to indicate that and it's necessary
> - * for next migration to work from this new version QEMU.
> - */
> - cs->gicd_no_migration_shift_bug = true;
> -
> - return 0;
> -}
> -
> static bool needed_always(void *opaque)
> {
> return true;
> @@ -210,7 +208,6 @@ const VMStateDescription vmstate_gicv3_gicd_no_migration_shift_bug = {
> .version_id = 1,
> .minimum_version_id = 1,
> .needed = needed_always,
> - .post_load = gicv3_gicd_no_migration_shift_bug_post_load,
> .fields = (VMStateField[]) {
> VMSTATE_BOOL(gicd_no_migration_shift_bug, GICv3State),
> VMSTATE_END_OF_LIST()
> --
> 2.17.1
>
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
next prev parent reply other threads:[~2018-08-06 15:39 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-08-06 12:34 [Qemu-arm] [PATCH for-3.0 v2 0/5] Arm migration fixes for 3.0 Peter Maydell
2018-08-06 12:34 ` [Qemu-devel] " Peter Maydell
2018-08-06 12:34 ` [Qemu-arm] [PATCH for-3.0 v2 1/5] hw/intc/arm_gicv3_common: Give no-migration-shift-bug subsection a needed function Peter Maydell
2018-08-06 12:34 ` [Qemu-devel] " Peter Maydell
2018-08-07 14:34 ` [Qemu-arm] " Juan Quintela
2018-08-07 14:34 ` [Qemu-devel] " Juan Quintela
2018-08-06 12:34 ` [Qemu-arm] [PATCH for-3.0 v2 2/5] hw/intc/arm_gicv3_common: Combine duplicate .subsections in vmstate_gicv3_cpu Peter Maydell
2018-08-06 12:34 ` [Qemu-devel] " Peter Maydell
2018-08-07 14:34 ` [Qemu-arm] " Juan Quintela
2018-08-07 14:34 ` [Qemu-devel] " Juan Quintela
2018-08-06 12:34 ` [Qemu-arm] [PATCH for-3.0 v2 3/5] target/arm: Add dummy needed functions to M profile vmstate subsections Peter Maydell
2018-08-06 12:34 ` [Qemu-devel] " Peter Maydell
2018-08-07 14:40 ` [Qemu-arm] " Juan Quintela
2018-08-07 14:40 ` [Qemu-devel] " Juan Quintela
2018-08-07 14:45 ` [Qemu-arm] " Peter Maydell
2018-08-07 14:45 ` [Qemu-devel] " Peter Maydell
2018-08-07 14:50 ` Juan Quintela
2018-08-07 14:50 ` Juan Quintela
2018-08-06 12:34 ` [Qemu-arm] [PATCH for-3.0 v2 4/5] hw/intc/arm_gicv3_common: Move post_load hooks to top-level VMSD Peter Maydell
2018-08-06 12:34 ` [Qemu-devel] " Peter Maydell
2018-08-06 15:32 ` [Qemu-arm] " Dr. David Alan Gilbert
2018-08-06 15:32 ` [Qemu-devel] " Dr. David Alan Gilbert
2018-08-07 14:41 ` [Qemu-arm] " Juan Quintela
2018-08-07 14:41 ` [Qemu-devel] " Juan Quintela
2018-08-06 12:34 ` [Qemu-arm] [PATCH for-3.0 v2 5/5] hw/intc/arm_gicv3_common: Move gicd shift bug handling to gicv3_post_load Peter Maydell
2018-08-06 12:34 ` [Qemu-devel] " Peter Maydell
2018-08-06 15:39 ` Dr. David Alan Gilbert [this message]
2018-08-06 15:39 ` Dr. David Alan Gilbert
2018-08-06 14:06 ` [Qemu-arm] [Qemu-devel] [PATCH for-3.0 v2 0/5] Arm migration fixes for 3.0 Richard Henderson
2018-08-06 14:06 ` Richard Henderson
2018-08-06 16:50 ` [Qemu-arm] " Peter Maydell
2018-08-06 16:50 ` Peter Maydell
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=20180806153925.GF2473@work-vm \
--to=dgilbert@redhat.com \
--cc=patches@linaro.org \
--cc=peter.maydell@linaro.org \
--cc=qemu-arm@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=shannon.zhaosl@gmail.com \
--cc=zhaoshenglong@huawei.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.