On 5/9/23 3:47 PM, Elliot Berman wrote:
Qualcomm platforms have a firmware entity which performs access control
to physical pages. Dynamically started Gunyah virtual machines use the
QCOM_SCM_RM_MANAGED_VMID for access. Linux thus needs to assign access
to the memory used by guest VMs. Gunyah doesn't do this operation for us
since it is the current VM (typically VMID_HLOS) delegating the access
and not Gunyah itself. Use the Gunyah platform ops to achieve this so
that only Qualcomm platforms attempt to make the needed SCM calls.
Co-developed-by: Prakruthi Deepak Heragu <quic_pheragu@xxxxxxxxxxx>
Signed-off-by: Prakruthi Deepak Heragu <quic_pheragu@xxxxxxxxxxx>
Signed-off-by: Elliot Berman <quic_eberman@xxxxxxxxxxx>
Minor suggestions below. Please consider them, but either way:
Reviewed-by: Alex Elder <elder@xxxxxxxxxx>
---
drivers/virt/gunyah/Kconfig | 13 +++
drivers/virt/gunyah/Makefile | 1 +
drivers/virt/gunyah/gunyah_qcom.c | 147 ++++++++++++++++++++++++++++++
3 files changed, 161 insertions(+)
create mode 100644 drivers/virt/gunyah/gunyah_qcom.c
diff --git a/drivers/virt/gunyah/Kconfig b/drivers/virt/gunyah/Kconfig
index de815189dab6..0421b751aad4 100644
--- a/drivers/virt/gunyah/Kconfig
+++ b/drivers/virt/gunyah/Kconfig
@@ -5,6 +5,7 @@ config GUNYAH
depends on ARM64
depends on MAILBOX
select GUNYAH_PLATFORM_HOOKS
+ imply GUNYAH_QCOM_PLATFORM if ARCH_QCOM
help
The Gunyah drivers are the helper interfaces that run in a guest VM
such as basic inter-VM IPC and signaling mechanisms, and higher level
@@ -15,3 +16,15 @@ config GUNYAH
config GUNYAH_PLATFORM_HOOKS
tristate
+
+config GUNYAH_QCOM_PLATFORM
+ tristate "Support for Gunyah on Qualcomm platforms"
+ depends on GUNYAH
+ select GUNYAH_PLATFORM_HOOKS
+ select QCOM_SCM
+ help
+ Enable support for interacting with Gunyah on Qualcomm
+ platforms. Interaction with Qualcomm firmware requires
+ extra platform-specific support.
+
+ Say Y/M here to use Gunyah on Qualcomm platforms.
diff --git a/drivers/virt/gunyah/Makefile b/drivers/virt/gunyah/Makefile
index 4fbeee521d60..2aa9ff038ed0 100644
--- a/drivers/virt/gunyah/Makefile
+++ b/drivers/virt/gunyah/Makefile
@@ -1,6 +1,7 @@
# SPDX-License-Identifier: GPL-2.0
obj-$(CONFIG_GUNYAH_PLATFORM_HOOKS) += gunyah_platform_hooks.o
+obj-$(CONFIG_GUNYAH_QCOM_PLATFORM) += gunyah_qcom.o
gunyah-y += rsc_mgr.o rsc_mgr_rpc.o vm_mgr.o vm_mgr_mm.o
obj-$(CONFIG_GUNYAH) += gunyah.o
diff --git a/drivers/virt/gunyah/gunyah_qcom.c b/drivers/virt/gunyah/gunyah_qcom.c
new file mode 100644
index 000000000000..18acbda8fcbd
--- /dev/null
+++ b/drivers/virt/gunyah/gunyah_qcom.c
@@ -0,0 +1,147 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (c) 2023 Qualcomm Innovation Center, Inc. All rights reserved.
+ */
+
+#include <linux/arm-smccc.h>
+#include <linux/gunyah_rsc_mgr.h>
+#include <linux/module.h>
+#include <linux/firmware/qcom/qcom_scm.h>
+#include <linux/types.h>
+#include <linux/uuid.h>
+
+#define QCOM_SCM_RM_MANAGED_VMID 0x3A
+#define QCOM_SCM_MAX_MANAGED_VMID 0x3F
Is this limited to 63 because there are at most 64 VMIDs
that can be represented in a 64-bit unsigned?
+
+static int qcom_scm_gh_rm_pre_mem_share(struct gh_rm *rm, struct gh_rm_mem_parcel *mem_parcel)
+{
+ struct qcom_scm_vmperm *new_perms;
+ u64 src, src_cpy;
+ int ret = 0, i, n;
+ u16 vmid;
+
+ new_perms = kcalloc(mem_parcel->n_acl_entries, sizeof(*new_perms), GFP_KERNEL);
+ if (!new_perms)
+ return -ENOMEM;
+
+ for (n = 0; n < mem_parcel->n_acl_entries; n++) {
+ vmid = le16_to_cpu(mem_parcel->acl_entries[n].vmid);
+ if (vmid <= QCOM_SCM_MAX_MANAGED_VMID)
+ new_perms[n].vmid = vmid;
+ else
+ new_perms[n].vmid = QCOM_SCM_RM_MANAGED_VMID;
So any out-of-range VM ID will cause the hunk of memory to
be assigned to the resource manager. Is it expected that
this can occur (and not be an error)?
+ if (mem_parcel->acl_entries[n].perms & GH_RM_ACL_X)
+ new_perms[n].perm |= QCOM_SCM_PERM_EXEC;
+ if (mem_parcel->acl_entries[n].perms & GH_RM_ACL_W)
+ new_perms[n].perm |= QCOM_SCM_PERM_WRITE;
+ if (mem_parcel->acl_entries[n].perms & GH_RM_ACL_R)
+ new_perms[n].perm |= QCOM_SCM_PERM_READ;
+ }
+
+ src = (1ull << QCOM_SCM_VMID_HLOS);
src = BIT_ULL(QCOM_SCM_VMID_HLOS);
+
+ for (i = 0; i < mem_parcel->n_mem_entries; i++) {
+ src_cpy = src;
+ ret = qcom_scm_assign_mem(le64_to_cpu(mem_parcel->mem_entries[i].phys_addr),
+ le64_to_cpu(mem_parcel->mem_entries[i].size),
+ &src_cpy, new_perms, mem_parcel->n_acl_entries);
Loops like this can look simpler if you jump to error handling
at the end that does this unwind activity, rather than incorporating
it inside the loop itself. Or even just breaking if ret != 0, e.g.:
if (ret)
break;
}
if (!ret)
return 0;
/* And do the following block here, "outdented" twice */
+ if (ret) {
+ src = 0;
+ for (n = 0; n < mem_parcel->n_acl_entries; n++) {
+ vmid = le16_to_cpu(mem_parcel->acl_entries[n].vmid);
+ if (vmid <= QCOM_SCM_MAX_MANAGED_VMID)
+ src |= (1ull << vmid);
src |= BIT_ULL(vmid);
+ else
+ src |= (1ull << QCOM_SCM_RM_MANAGED_VMID);
src |= BIT_ULL(QCOM_SCM_RM_MANAGED_VMID);
+ }
+
+ new_perms[0].vmid = QCOM_SCM_VMID_HLOS;
+
+ for (i--; i >= 0; i--) {
+ src_cpy = src;
+ WARN_ON_ONCE(qcom_scm_assign_mem(
+ le64_to_cpu(mem_parcel->mem_entries[i].phys_addr),
+ le64_to_cpu(mem_parcel->mem_entries[i].size),
+ &src_cpy, new_perms, 1));
+ }
+ break;
+ }
+ }
+
+ kfree(new_perms);
+ return ret;
+}
+
+static int qcom_scm_gh_rm_post_mem_reclaim(struct gh_rm *rm, struct gh_rm_mem_parcel *mem_parcel)
+{
+ struct qcom_scm_vmperm new_perms;
+ u64 src = 0, src_cpy;
+ int ret = 0, i, n;
+ u16 vmid;
+
+ new_perms.vmid = QCOM_SCM_VMID_HLOS;
+ new_perms.perm = QCOM_SCM_PERM_EXEC | QCOM_SCM_PERM_WRITE | QCOM_SCM_PERM_READ;
+
+ for (n = 0; n < mem_parcel->n_acl_entries; n++) {
+ vmid = le16_to_cpu(mem_parcel->acl_entries[n].vmid);
+ if (vmid <= QCOM_SCM_MAX_MANAGED_VMID)
+ src |= (1ull << vmid);
+ else
+ src |= (1ull << QCOM_SCM_RM_MANAGED_VMID);
+ }
+
+ for (i = 0; i < mem_parcel->n_mem_entries; i++) {
+ src_cpy = src;
+ ret = qcom_scm_assign_mem(le64_to_cpu(mem_parcel->mem_entries[i].phys_addr),
+ le64_to_cpu(mem_parcel->mem_entries[i].size),
+ &src_cpy, &new_perms, 1);
+ WARN_ON_ONCE(ret);
+ }
+
+ return ret;
+}
+
+static struct gh_rm_platform_ops qcom_scm_gh_rm_platform_ops = {
+ .pre_mem_share = qcom_scm_gh_rm_pre_mem_share,
+ .post_mem_reclaim = qcom_scm_gh_rm_post_mem_reclaim,
+};
+
+/* {19bd54bd-0b37-571b-946f-609b54539de6} */
+static const uuid_t QCOM_EXT_UUID =
+ UUID_INIT(0x19bd54bd, 0x0b37, 0x571b, 0x94, 0x6f, 0x60, 0x9b, 0x54, 0x53, 0x9d, 0xe6);
+
+#define GH_QCOM_EXT_CALL_UUID_ID ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, ARM_SMCCC_SMC_32, \
+ ARM_SMCCC_OWNER_VENDOR_HYP, 0x3f01)
+
+static bool gh_has_qcom_extensions(void)
+{
+ struct arm_smccc_res res;
+ uuid_t uuid;
+
+ arm_smccc_1_1_smc(GH_QCOM_EXT_CALL_UUID_ID, &res);
+
+ ((u32 *)&uuid.b[0])[0] = lower_32_bits(res.a0);
+ ((u32 *)&uuid.b[0])[1] = lower_32_bits(res.a1);
+ ((u32 *)&uuid.b[0])[2] = lower_32_bits(res.a2);
+ ((u32 *)&uuid.b[0])[3] = lower_32_bits(res.a3);
I said this elsewhere. I'd rather see:
u32 *u = (u32 *)&uuid; /* Or &uuid.b? */
*u++ = lower_32_bits(res.a0);
. . .
+
+ return uuid_equal(&uuid, &QCOM_EXT_UUID);
+}
+
+static int __init qcom_gh_platform_hooks_register(void)
+{
+ if (!gh_has_qcom_extensions())
+ return -ENODEV;
+
+ return gh_rm_register_platform_ops(&qcom_scm_gh_rm_platform_ops);
+}
+
+static void __exit qcom_gh_platform_hooks_unregister(void)
+{
+ gh_rm_unregister_platform_ops(&qcom_scm_gh_rm_platform_ops);
+}
+
+module_init(qcom_gh_platform_hooks_register);
+module_exit(qcom_gh_platform_hooks_unregister);
+MODULE_DESCRIPTION("Qualcomm Technologies, Inc. Platform Hooks for Gunyah");
+MODULE_LICENSE("GPL");