* [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT
@ 2025-08-15 2:26 Bibo Mao
2025-08-15 2:26 ` [PATCH v3 1/4] LoongArch: KVM: Fix stack protector issue in send_ipi_data() Bibo Mao
` (4 more replies)
0 siblings, 5 replies; 7+ messages in thread
From: Bibo Mao @ 2025-08-15 2:26 UTC (permalink / raw)
To: Tianrui Zhao, Huacai Chen, Xianglai Li; +Cc: kvm, loongarch, linux-kernel
Thre are some small enhancement about IPI emulation and LBT enabling in
LoongArch KVM. With IPI, it supports sending command to vCPU itself. And
with LBT it adds flag checking int function kvm_own_lbt() and make it
robust.
---
v2 ... v3:
1. Fix stack protector issue in send_ipi_data()
v1 ... v2:
1. Add sending IPI command to vCPU itself
2. Avoid duplicated LBT enabling in kvm_own_lbt()
---
Bibo Mao (4):
LoongArch: KVM: Fix stack protector issue in send_ipi_data()
LoongArch: KVM: Access mailbox directly in mail_send()
LoongArch: KVM: Add implementation with IOCSR_IPI_SET
LoongArch: KVM: Make function kvm_own_lbt() robust
arch/loongarch/kvm/intc/ipi.c | 57 ++++++++++++++++++++++-------------
arch/loongarch/kvm/vcpu.c | 8 +++--
2 files changed, 41 insertions(+), 24 deletions(-)
base-commit: dfc0f6373094dd88e1eaf76c44f2ff01b65db851
--
2.39.3
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH v3 1/4] LoongArch: KVM: Fix stack protector issue in send_ipi_data()
2025-08-15 2:26 [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Bibo Mao
@ 2025-08-15 2:26 ` Bibo Mao
2025-08-15 2:26 ` [PATCH v3 2/4] LoongArch: KVM: Access mailbox directly in mail_send() Bibo Mao
` (3 subsequent siblings)
4 siblings, 0 replies; 7+ messages in thread
From: Bibo Mao @ 2025-08-15 2:26 UTC (permalink / raw)
To: Tianrui Zhao, Huacai Chen, Xianglai Li
Cc: kvm, loongarch, linux-kernel, stable
Function kvm_io_bus_read() is called in function send_ipi_data(), buffer
size of parameter *val should be at least 8 bytes. Some emulation
functions like loongarch_ipi_readl() and kvm_eiointc_read() will
write buffer *val with 8 bytes signed extension regardless parameter
len.
Otherwise there will be buffer overflow issue when CONFIG_STACKPROTECTOR
is enabled. The bug report is shown as follows:
Kernel panic - not syncing: stack-protector: Kernel stack is corrupted in: send_ipi_data+0x194/0x1a0 [kvm]
CPU: 11 UID: 107 PID: 2692 Comm: CPU 0/KVM Not tainted 6.17.0-rc1+ #102 PREEMPT(full)
Stack : 9000000005901568 0000000000000000 9000000003af371c 900000013c68c000
900000013c68f850 900000013c68f858 0000000000000000 900000013c68f998
900000013c68f990 900000013c68f990 900000013c68f6c0 fffffffffffdb058
fffffffffffdb0e0 900000013c68f858 911e1d4d39cf0ec2 9000000105657a00
0000000000000001 fffffffffffffffe 0000000000000578 282049464555206e
6f73676e6f6f4c20 0000000000000001 00000000086b4000 0000000000000000
0000000000000000 0000000000000000 9000000005709968 90000000058f9000
900000013c68fa68 900000013c68fab4 90000000029279f0 900000010153f940
900000010001f360 0000000000000000 9000000003af3734 000000004390000c
00000000000000b0 0000000000000004 0000000000000000 0000000000071c1d
...
Call Trace:
[<9000000003af3734>] show_stack+0x5c/0x180
[<9000000003aed168>] dump_stack_lvl+0x6c/0x9c
[<9000000003ad0ab0>] vpanic+0x108/0x2c4
[<9000000003ad0ca8>] panic+0x3c/0x40
[<9000000004eb0a1c>] __stack_chk_fail+0x14/0x18
[<ffff8000023473f8>] send_ipi_data+0x190/0x1a0 [kvm]
[<ffff8000023313e4>] __kvm_io_bus_write+0xa4/0xe8 [kvm]
[<ffff80000233147c>] kvm_io_bus_write+0x54/0x90 [kvm]
[<ffff80000233f9f8>] kvm_emu_iocsr+0x180/0x310 [kvm]
[<ffff80000233fe08>] kvm_handle_gspr+0x280/0x478 [kvm]
[<ffff8000023443e8>] kvm_handle_exit+0xc0/0x130 [kvm]
Fixes: daee2f9cae551 ("LoongArch: KVM: Add IPI read and write function")
Cc: stable@vger.kernel.org
Signed-off-by: Bibo Mao <maobibo@loongson.cn>
---
arch/loongarch/kvm/intc/ipi.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/loongarch/kvm/intc/ipi.c b/arch/loongarch/kvm/intc/ipi.c
index e658d5b37c04..7925651d2ccf 100644
--- a/arch/loongarch/kvm/intc/ipi.c
+++ b/arch/loongarch/kvm/intc/ipi.c
@@ -99,7 +99,7 @@ static void write_mailbox(struct kvm_vcpu *vcpu, int offset, uint64_t data, int
static int send_ipi_data(struct kvm_vcpu *vcpu, gpa_t addr, uint64_t data)
{
int i, idx, ret;
- uint32_t val = 0, mask = 0;
+ uint64_t val = 0, mask = 0;
/*
* Bit 27-30 is mask for byte writing.
@@ -108,7 +108,7 @@ static int send_ipi_data(struct kvm_vcpu *vcpu, gpa_t addr, uint64_t data)
if ((data >> 27) & 0xf) {
/* Read the old val */
idx = srcu_read_lock(&vcpu->kvm->srcu);
- ret = kvm_io_bus_read(vcpu, KVM_IOCSR_BUS, addr, sizeof(val), &val);
+ ret = kvm_io_bus_read(vcpu, KVM_IOCSR_BUS, addr, 4, &val);
srcu_read_unlock(&vcpu->kvm->srcu, idx);
if (unlikely(ret)) {
kvm_err("%s: : read data from addr %llx failed\n", __func__, addr);
@@ -124,7 +124,7 @@ static int send_ipi_data(struct kvm_vcpu *vcpu, gpa_t addr, uint64_t data)
}
val |= ((uint32_t)(data >> 32) & ~mask);
idx = srcu_read_lock(&vcpu->kvm->srcu);
- ret = kvm_io_bus_write(vcpu, KVM_IOCSR_BUS, addr, sizeof(val), &val);
+ ret = kvm_io_bus_write(vcpu, KVM_IOCSR_BUS, addr, 4, &val);
srcu_read_unlock(&vcpu->kvm->srcu, idx);
if (unlikely(ret))
kvm_err("%s: : write data to addr %llx failed\n", __func__, addr);
--
2.39.3
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v3 2/4] LoongArch: KVM: Access mailbox directly in mail_send()
2025-08-15 2:26 [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Bibo Mao
2025-08-15 2:26 ` [PATCH v3 1/4] LoongArch: KVM: Fix stack protector issue in send_ipi_data() Bibo Mao
@ 2025-08-15 2:26 ` Bibo Mao
2025-08-15 2:26 ` [PATCH v3 3/4] LoongArch: KVM: Add implementation with IOCSR_IPI_SET Bibo Mao
` (2 subsequent siblings)
4 siblings, 0 replies; 7+ messages in thread
From: Bibo Mao @ 2025-08-15 2:26 UTC (permalink / raw)
To: Tianrui Zhao, Huacai Chen, Xianglai Li; +Cc: kvm, loongarch, linux-kernel
With function mail_send(), it is to write mailbox of other VCPUs.
Existing simple APIs read_mailbox/write_mailbox can be used directly
rather than send command on IOCSR address.
Signed-off-by: Bibo Mao <maobibo@loongson.cn>
---
arch/loongarch/kvm/intc/ipi.c | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/arch/loongarch/kvm/intc/ipi.c b/arch/loongarch/kvm/intc/ipi.c
index 7925651d2ccf..f27f79a0c1e0 100644
--- a/arch/loongarch/kvm/intc/ipi.c
+++ b/arch/loongarch/kvm/intc/ipi.c
@@ -134,7 +134,8 @@ static int send_ipi_data(struct kvm_vcpu *vcpu, gpa_t addr, uint64_t data)
static int mail_send(struct kvm *kvm, uint64_t data)
{
- int cpu, mailbox, offset;
+ int i, cpu, mailbox, offset;
+ uint32_t val = 0, mask = 0;
struct kvm_vcpu *vcpu;
cpu = ((data & 0xffffffff) >> 16) & 0x3ff;
@@ -144,9 +145,18 @@ static int mail_send(struct kvm *kvm, uint64_t data)
return -EINVAL;
}
mailbox = ((data & 0xffffffff) >> 2) & 0x7;
- offset = IOCSR_IPI_BASE + IOCSR_IPI_BUF_20 + mailbox * 4;
+ offset = IOCSR_IPI_BUF_20 + mailbox * 4;
+ if ((data >> 27) & 0xf) {
+ val = read_mailbox(vcpu, offset, 4);
+ for (i = 0; i < 4; i++)
+ if (data & (BIT(27 + i)))
+ mask |= (0xff << (i * 8));
+ val &= mask;
+ }
- return send_ipi_data(vcpu, offset, data);
+ val |= ((uint32_t)(data >> 32) & ~mask);
+ write_mailbox(vcpu, offset, val, 4);
+ return 0;
}
static int any_send(struct kvm *kvm, uint64_t data)
--
2.39.3
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v3 3/4] LoongArch: KVM: Add implementation with IOCSR_IPI_SET
2025-08-15 2:26 [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Bibo Mao
2025-08-15 2:26 ` [PATCH v3 1/4] LoongArch: KVM: Fix stack protector issue in send_ipi_data() Bibo Mao
2025-08-15 2:26 ` [PATCH v3 2/4] LoongArch: KVM: Access mailbox directly in mail_send() Bibo Mao
@ 2025-08-15 2:26 ` Bibo Mao
2025-08-15 2:26 ` [PATCH v3 4/4] LoongArch: KVM: Make function kvm_own_lbt() robust Bibo Mao
2025-08-17 3:45 ` [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Huacai Chen
4 siblings, 0 replies; 7+ messages in thread
From: Bibo Mao @ 2025-08-15 2:26 UTC (permalink / raw)
To: Tianrui Zhao, Huacai Chen, Xianglai Li; +Cc: kvm, loongarch, linux-kernel
IPI IOCSR register IOCSR_IPI_SET can send ipi interrupt to other vCPUs,
also it can send interrupt to vCPU itself. Instead there is such
operation on Linux such as arch_irq_work_raise(), it will send ipi
message to vCPU itself.
Here add implementation of write operation with IOCSR_IPI_SET register.
Signed-off-by: Bibo Mao <maobibo@loongson.cn>
---
arch/loongarch/kvm/intc/ipi.c | 35 ++++++++++++++++++++---------------
1 file changed, 20 insertions(+), 15 deletions(-)
diff --git a/arch/loongarch/kvm/intc/ipi.c b/arch/loongarch/kvm/intc/ipi.c
index f27f79a0c1e0..3fb98abd9554 100644
--- a/arch/loongarch/kvm/intc/ipi.c
+++ b/arch/loongarch/kvm/intc/ipi.c
@@ -7,13 +7,26 @@
#include <asm/kvm_ipi.h>
#include <asm/kvm_vcpu.h>
-static void ipi_send(struct kvm *kvm, uint64_t data)
+static void ipi_set(struct kvm_vcpu *vcpu, uint32_t data)
{
- int cpu, action;
uint32_t status;
- struct kvm_vcpu *vcpu;
struct kvm_interrupt irq;
+ spin_lock(&vcpu->arch.ipi_state.lock);
+ status = vcpu->arch.ipi_state.status;
+ vcpu->arch.ipi_state.status |= data;
+ spin_unlock(&vcpu->arch.ipi_state.lock);
+ if ((status == 0) && data) {
+ irq.irq = LARCH_INT_IPI;
+ kvm_vcpu_ioctl_interrupt(vcpu, &irq);
+ }
+}
+
+static void ipi_send(struct kvm *kvm, uint64_t data)
+{
+ int cpu;
+ struct kvm_vcpu *vcpu;
+
cpu = ((data & 0xffffffff) >> 16) & 0x3ff;
vcpu = kvm_get_vcpu_by_cpuid(kvm, cpu);
if (unlikely(vcpu == NULL)) {
@@ -21,15 +34,7 @@ static void ipi_send(struct kvm *kvm, uint64_t data)
return;
}
- action = BIT(data & 0x1f);
- spin_lock(&vcpu->arch.ipi_state.lock);
- status = vcpu->arch.ipi_state.status;
- vcpu->arch.ipi_state.status |= action;
- spin_unlock(&vcpu->arch.ipi_state.lock);
- if (status == 0) {
- irq.irq = LARCH_INT_IPI;
- kvm_vcpu_ioctl_interrupt(vcpu, &irq);
- }
+ ipi_set(vcpu, BIT(data & 0x1f));
}
static void ipi_clear(struct kvm_vcpu *vcpu, uint64_t data)
@@ -241,7 +246,7 @@ static int loongarch_ipi_writel(struct kvm_vcpu *vcpu, gpa_t addr, int len, cons
spin_unlock(&vcpu->arch.ipi_state.lock);
break;
case IOCSR_IPI_SET:
- ret = -EINVAL;
+ ipi_set(vcpu, data);
break;
case IOCSR_IPI_CLEAR:
/* Just clear the status of the current vcpu */
@@ -260,10 +265,10 @@ static int loongarch_ipi_writel(struct kvm_vcpu *vcpu, gpa_t addr, int len, cons
ipi_send(vcpu->kvm, data);
break;
case IOCSR_MAIL_SEND:
- ret = mail_send(vcpu->kvm, *(uint64_t *)val);
+ ret = mail_send(vcpu->kvm, data);
break;
case IOCSR_ANY_SEND:
- ret = any_send(vcpu->kvm, *(uint64_t *)val);
+ ret = any_send(vcpu->kvm, data);
break;
default:
kvm_err("%s: unknown addr: %llx\n", __func__, addr);
--
2.39.3
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v3 4/4] LoongArch: KVM: Make function kvm_own_lbt() robust
2025-08-15 2:26 [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Bibo Mao
` (2 preceding siblings ...)
2025-08-15 2:26 ` [PATCH v3 3/4] LoongArch: KVM: Add implementation with IOCSR_IPI_SET Bibo Mao
@ 2025-08-15 2:26 ` Bibo Mao
2025-08-17 3:45 ` [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Huacai Chen
4 siblings, 0 replies; 7+ messages in thread
From: Bibo Mao @ 2025-08-15 2:26 UTC (permalink / raw)
To: Tianrui Zhao, Huacai Chen, Xianglai Li; +Cc: kvm, loongarch, linux-kernel
Add flag KVM_LARCH_LBT checking in function kvm_own_lbt(), so that
it can be called safely rather than duplicated enabling again.
Signed-off-by: Bibo Mao <maobibo@loongson.cn>
---
arch/loongarch/kvm/vcpu.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/arch/loongarch/kvm/vcpu.c b/arch/loongarch/kvm/vcpu.c
index d1b8c50941ca..ce478151466c 100644
--- a/arch/loongarch/kvm/vcpu.c
+++ b/arch/loongarch/kvm/vcpu.c
@@ -1283,9 +1283,11 @@ int kvm_own_lbt(struct kvm_vcpu *vcpu)
return -EINVAL;
preempt_disable();
- set_csr_euen(CSR_EUEN_LBTEN);
- _restore_lbt(&vcpu->arch.lbt);
- vcpu->arch.aux_inuse |= KVM_LARCH_LBT;
+ if (!(vcpu->arch.aux_inuse & KVM_LARCH_LBT)) {
+ set_csr_euen(CSR_EUEN_LBTEN);
+ _restore_lbt(&vcpu->arch.lbt);
+ vcpu->arch.aux_inuse |= KVM_LARCH_LBT;
+ }
preempt_enable();
return 0;
--
2.39.3
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT
2025-08-15 2:26 [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Bibo Mao
` (3 preceding siblings ...)
2025-08-15 2:26 ` [PATCH v3 4/4] LoongArch: KVM: Make function kvm_own_lbt() robust Bibo Mao
@ 2025-08-17 3:45 ` Huacai Chen
2025-08-29 10:11 ` Huacai Chen
4 siblings, 1 reply; 7+ messages in thread
From: Huacai Chen @ 2025-08-17 3:45 UTC (permalink / raw)
To: Bibo Mao; +Cc: Tianrui Zhao, Xianglai Li, kvm, loongarch, linux-kernel
Patch-1 and Patch-4 are applied for loongarch-fixes.
Huacai
On Fri, Aug 15, 2025 at 10:26 AM Bibo Mao <maobibo@loongson.cn> wrote:
>
> Thre are some small enhancement about IPI emulation and LBT enabling in
> LoongArch KVM. With IPI, it supports sending command to vCPU itself. And
> with LBT it adds flag checking int function kvm_own_lbt() and make it
> robust.
>
> ---
> v2 ... v3:
> 1. Fix stack protector issue in send_ipi_data()
>
> v1 ... v2:
> 1. Add sending IPI command to vCPU itself
> 2. Avoid duplicated LBT enabling in kvm_own_lbt()
> ---
> Bibo Mao (4):
> LoongArch: KVM: Fix stack protector issue in send_ipi_data()
> LoongArch: KVM: Access mailbox directly in mail_send()
> LoongArch: KVM: Add implementation with IOCSR_IPI_SET
> LoongArch: KVM: Make function kvm_own_lbt() robust
>
> arch/loongarch/kvm/intc/ipi.c | 57 ++++++++++++++++++++++-------------
> arch/loongarch/kvm/vcpu.c | 8 +++--
> 2 files changed, 41 insertions(+), 24 deletions(-)
>
>
> base-commit: dfc0f6373094dd88e1eaf76c44f2ff01b65db851
> --
> 2.39.3
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT
2025-08-17 3:45 ` [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Huacai Chen
@ 2025-08-29 10:11 ` Huacai Chen
0 siblings, 0 replies; 7+ messages in thread
From: Huacai Chen @ 2025-08-29 10:11 UTC (permalink / raw)
To: Bibo Mao; +Cc: Tianrui Zhao, Xianglai Li, kvm, loongarch, linux-kernel
On Sun, Aug 17, 2025 at 11:45 AM Huacai Chen <chenhuacai@kernel.org> wrote:
>
> Patch-1 and Patch-4 are applied for loongarch-fixes.
The rest are applied for loongarch-kvm.
Huacai
>
> Huacai
>
> On Fri, Aug 15, 2025 at 10:26 AM Bibo Mao <maobibo@loongson.cn> wrote:
> >
> > Thre are some small enhancement about IPI emulation and LBT enabling in
> > LoongArch KVM. With IPI, it supports sending command to vCPU itself. And
> > with LBT it adds flag checking int function kvm_own_lbt() and make it
> > robust.
> >
> > ---
> > v2 ... v3:
> > 1. Fix stack protector issue in send_ipi_data()
> >
> > v1 ... v2:
> > 1. Add sending IPI command to vCPU itself
> > 2. Avoid duplicated LBT enabling in kvm_own_lbt()
> > ---
> > Bibo Mao (4):
> > LoongArch: KVM: Fix stack protector issue in send_ipi_data()
> > LoongArch: KVM: Access mailbox directly in mail_send()
> > LoongArch: KVM: Add implementation with IOCSR_IPI_SET
> > LoongArch: KVM: Make function kvm_own_lbt() robust
> >
> > arch/loongarch/kvm/intc/ipi.c | 57 ++++++++++++++++++++++-------------
> > arch/loongarch/kvm/vcpu.c | 8 +++--
> > 2 files changed, 41 insertions(+), 24 deletions(-)
> >
> >
> > base-commit: dfc0f6373094dd88e1eaf76c44f2ff01b65db851
> > --
> > 2.39.3
> >
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2025-08-29 10:11 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-08-15 2:26 [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Bibo Mao
2025-08-15 2:26 ` [PATCH v3 1/4] LoongArch: KVM: Fix stack protector issue in send_ipi_data() Bibo Mao
2025-08-15 2:26 ` [PATCH v3 2/4] LoongArch: KVM: Access mailbox directly in mail_send() Bibo Mao
2025-08-15 2:26 ` [PATCH v3 3/4] LoongArch: KVM: Add implementation with IOCSR_IPI_SET Bibo Mao
2025-08-15 2:26 ` [PATCH v3 4/4] LoongArch: KVM: Make function kvm_own_lbt() robust Bibo Mao
2025-08-17 3:45 ` [PATCH v3 0/4] LoongArch: KVM: Small enhancements about IPI and LBT Huacai Chen
2025-08-29 10:11 ` Huacai Chen
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).