Skip to contentSkip to conversation

s390x: initial support for virtio-mem

RFC PATCH Superseded David Hildenbrand 4 participants 39 messages Jul 8, 2020 – Jul 20, 2020

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

cover s390x: initial support for virtio-mem

This wires up the initial, basic version of virito-mem for s390x. General information about virtio-mem can be found at [1] and in QEMU commit [2]. Patch #5 contains a short example for s390x.

virtio-mem for x86-64 Linux is part of v5.8-rc1. A branch with a s390x prototype can be found at:

    [email protected]:davidhildenbrand/linux.git virtio-mem-s390x

Note that the kernel should either be compiled via
CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE, or "memhp_default_state=online" should be passed on the kernel cmdline.

This series can be found at:

    [email protected]:davidhildenbrand/qemu.git virtio-mem-s390x

Related to s390x, we'll have to tackle migration of storage keys and storage attributes (especially, skipping unplugged parts). Not sure if I am missing something else (any ideas?). For virtio-mem in general, there are a couple of TODOs, e.g., documented in [1] and [2], both in QEMU and Linux. However, the basics are around.

I only tested this with fairly small amount of RAM in a z/VM environemnt
...

[1] https://virtio-mem.gitlab.io/
[2] 910b25766b33 ("virtio-mem: Paravirtualized memory hot(un)plug")

David Hildenbrand (5):
s390x: move setting of maximum ram size to machine init
s390x: implement diag260
s390x: prepare device memory address space
s390x: implement virtio-mem-ccw
s390x: initial support for virtio-mem

 hw/s390x/Kconfig                   |   1 +
 hw/s390x/Makefile.objs             |   1 +
 hw/s390x/s390-virtio-ccw.c         | 178 ++++++++++++++++++++++++++++-
 hw/s390x/sclp.c                    |  32 ++----
 hw/s390x/virtio-ccw-mem.c          | 165 ++++++++++++++++++++++++++
 hw/s390x/virtio-ccw.h              |  13 +++
 hw/virtio/virtio-mem.c             |   2 +
 include/hw/s390x/s390-virtio-ccw.h |   3 +
 target/s390x/diag.c                |  57 +++++++++
 target/s390x/internal.h            |   2 +
 target/s390x/kvm.c                 |  11 ++
 target/s390x/misc_helper.c         |   6 +
 target/s390x/translate.c           |   4 +
 13 files changed, 449 insertions(+), 26 deletions(-)
 create mode 100644 hw/s390x/virtio-ccw-mem.c
Signature · 1 line
2.26.2
#1 lore raw

RFC PATCH 1/5 s390x: move setting of maximum ram size to machine init

As we no longer fixup the maximum ram size in sclp code, let's move setting the maximum ram size to ccw_init()->s390_memory_init(), which now looks like a better fit.

---
 hw/s390x/s390-virtio-ccw.c | 19 ++++++++++++++++---
 hw/s390x/sclp.c            | 20 +-------------------
 2 files changed, 17 insertions(+), 22 deletions(-)
hw/s390x/s390-virtio-ccw.c +16 -3
index 023fd25f2b..2e6d292c23 100644
@@ -160,13 +160,26 @@ static void virtio_ccw_register_hcalls(void)
                                    virtio_ccw_hcall_early_printk);
 }
 
-static void s390_memory_init(MemoryRegion *ram)
+static void s390_memory_init(MachineState *machine)
 {
     MemoryRegion *sysmem = get_system_memory();
     Error *local_err = NULL;
+    uint64_t hw_limit;
+    int ret;
+
+    /* We have to set the memory limit before adding any regions to sysmem. */
+    ret = s390_set_memory_limit(machine->maxram_size, &hw_limit);
+    if (ret == -E2BIG) {
+        error_report("host supports a maximum of %" PRIu64 " GB",
+                     hw_limit / GiB);
+        exit(EXIT_FAILURE);
+    } else if (ret) {
+        error_report("setting the guest size failed");
+        exit(EXIT_FAILURE);
+    }
 
     /* allocate RAM for core */
-    memory_region_add_subregion(sysmem, 0, ram);
+    memory_region_add_subregion(sysmem, 0, machine->ram);
 
     /*
      * Configure the maximum page size. As no memory devices were created
@@ -249,7 +262,7 @@ static void ccw_init(MachineState *machine)
 
     s390_sclp_init();
     /* init memory + setup max page size. Required for the CPU model */
-    s390_memory_init(machine->ram);
+    s390_memory_init(machine);
 
     /* init CPUs (incl. CPU model) early so s390_has_feature() works */
     s390_init_cpus(machine);
hw/s390x/sclp.c +1 -19
index d39f6d7785..f59195e15a 100644
@@ -327,32 +327,14 @@ void s390_sclp_init(void)
 
 static void sclp_realize(DeviceState *dev, Error **errp)
 {
-    MachineState *machine = MACHINE(qdev_get_machine());
     SCLPDevice *sclp = SCLP(dev);
-    Error *err = NULL;
-    uint64_t hw_limit;
-    int ret;
 
     /*
      * qdev_device_add searches the sysbus for TYPE_SCLP_EVENTS_BUS. As long
      * as we can't find a fitting bus via the qom tree, we have to add the
      * event facility to the sysbus, so e.g. a sclp console can be created.
      */
-    sysbus_realize(SYS_BUS_DEVICE(sclp->event_facility), &err);
-    if (err) {
-        goto out;
-    }
-
-    ret = s390_set_memory_limit(machine->maxram_size, &hw_limit);
-    if (ret == -E2BIG) {
-        error_setg(&err, "host supports a maximum of %" PRIu64 " GB",
-                   hw_limit / GiB);
-    } else if (ret) {
-        error_setg(&err, "setting the guest size failed");
-    }
-
-out:
-    error_propagate(errp, err);
+    sysbus_realize(SYS_BUS_DEVICE(sclp->event_facility), errp);
 }
 
 static void sclp_memory_init(SCLPDevice *sclp)
Signature · 1 line
2.26.2
#2 lore raw

RFC PATCH 2/5 s390x: implement diag260

In reply to David Hildenbrand

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

---
 target/s390x/diag.c        | 57 ++++++++++++++++++++++++++++++++++++++
 target/s390x/internal.h    |  2 ++
 target/s390x/kvm.c         | 11 ++++++++
 target/s390x/misc_helper.c |  6 ++++
 target/s390x/translate.c   |  4 +++
 5 files changed, 80 insertions(+)
target/s390x/diag.c +57 -0
index 1a48429564..c3b1e24b2c 100644
@@ -23,6 +23,63 @@
 #include "hw/s390x/pv.h"
 #include "kvm_s390x.h"
 
+void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
+{
+    MachineState *ms = MACHINE(qdev_get_machine());
+    const ram_addr_t initial_ram_size = ms->ram_size;
+    const uint64_t subcode = env->regs[r3];
+    S390CPU *cpu = env_archcpu(env);
+    ram_addr_t addr, length;
+    uint64_t tmp;
+
+    /* TODO: Unlock with new QEMU machine. */
+    if (false) {
+        s390_program_interrupt(env, PGM_OPERATION, ra);
+        return;
+    }
+
+    /*
+     * There also seems to be subcode "0xc", which stores the size of the
+     * first chunk and the total size to r1/r2. It's only used by very old
+     * Linux, so don't implement it.
+     */
+    if ((r1 & 1) || subcode != 0x10) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+    addr = env->regs[r1];
+    length = env->regs[r1 + 1];
+
+    /* FIXME: Somebody with documentation should fix this. */
+    if (!QEMU_IS_ALIGNED(addr, 16) || !QEMU_IS_ALIGNED(length, 16)) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */
+    if (!length) {
+        setcc(cpu, 3);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */
+    if (!address_space_access_valid(&address_space_memory, addr, length, true,
+                                    MEMTXATTRS_UNSPECIFIED)) {
+        s390_program_interrupt(env, PGM_ADDRESSING, ra);
+        return;
+    }
+
+    /* Indicate our initial memory ([0 .. ram_size - 1]) */
+    tmp = cpu_to_be64(0);
+    cpu_physical_memory_write(addr, &tmp, sizeof(tmp));
+    tmp = cpu_to_be64(initial_ram_size - 1);
+    cpu_physical_memory_write(addr + sizeof(tmp), &tmp, sizeof(tmp));
+
+    /* Exactly one entry was stored. */
+    env->regs[r3] = 1;
+    setcc(cpu, 0);
+}
+
 int handle_diag_288(CPUS390XState *env, uint64_t r1, uint64_t r3)
 {
     uint64_t func = env->regs[r1];
target/s390x/internal.h +2 -0
index b1e0ebf67f..a7a3df9a3b 100644
@@ -372,6 +372,8 @@ int mmu_translate_real(CPUS390XState *env, target_ulong raddr, int rw,
 
 
 /* misc_helper.c */
+void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3,
+                     uintptr_t ra);
 int handle_diag_288(CPUS390XState *env, uint64_t r1, uint64_t r3);
 void handle_diag_308(CPUS390XState *env, uint64_t r1, uint64_t r3,
                      uintptr_t ra);
target/s390x/kvm.c +11 -0
index f2f75d2a57..d6de3ad86c 100644
@@ -1565,6 +1565,14 @@ static int handle_hypercall(S390CPU *cpu, struct kvm_run *run)
     return ret;
 }
 
+static void kvm_handle_diag_260(S390CPU *cpu, struct kvm_run *run)
+{
+    const uint64_t r1 = (run->s390_sieic.ipa & 0x00f0) >> 4;
+    const uint64_t r3 = run->s390_sieic.ipa & 0x000f;
+
+    handle_diag_260(&cpu->env, r1, r3, 0);
+}
+
 static void kvm_handle_diag_288(S390CPU *cpu, struct kvm_run *run)
 {
     uint64_t r1, r3;
@@ -1614,6 +1622,9 @@ static int handle_diag(S390CPU *cpu, struct kvm_run *run, uint32_t ipb)
      */
     func_code = decode_basedisp_rs(&cpu->env, ipb, NULL) & DIAG_KVM_CODE_MASK;
     switch (func_code) {
+    case 0x260:
+        kvm_handle_diag_260(cpu, run);
+        break;
     case DIAG_TIMEREVENT:
         kvm_handle_diag_288(cpu, run);
         break;
target/s390x/misc_helper.c +6 -0
index 58dbc023eb..d7274eb320 100644
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();
target/s390x/translate.c +4 -0
index 4f6f1e31cd..6bb8b6e513 100644
@@ -2398,6 +2398,10 @@ static DisasJumpType op_diag(DisasContext *s, DisasOps *o)
     TCGv_i32 func_code = tcg_const_i32(get_field(s, i2));
 
     gen_helper_diag(cpu_env, r1, r3, func_code);
+    /* Only some diags modify the CC. */
+    if (get_field(s, i2) == 0x260) {
+        set_cc_static(s);
+    }
 
     tcg_temp_free_i32(func_code);
     tcg_temp_free_i32(r3);
Signature · 1 line
2.26.2
#3 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 46 lines · On Wed, 8 Jul 2020 20:51:32 +0200 David Hildenbrand <[email protected]> wrote:

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

---
 target/s390x/diag.c        | 57 ++++++++++++++++++++++++++++++++++++++
 target/s390x/internal.h    |  2 ++
 target/s390x/kvm.c         | 11 ++++++++
 target/s390x/misc_helper.c |  6 ++++
 target/s390x/translate.c   |  4 +++
 5 files changed, 80 insertions(+)
target/s390x/diag.c +9 -0
index 1a48429564..c3b1e24b2c 100644
@@ -23,6 +23,63 @@
 #include "hw/s390x/pv.h"
 #include "kvm_s390x.h"
 
+void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
+{
+    MachineState *ms = MACHINE(qdev_get_machine());
+    const ram_addr_t initial_ram_size = ms->ram_size;
+    const uint64_t subcode = env->regs[r3];
+    S390CPU *cpu = env_archcpu(env);
+    ram_addr_t addr, length;
+    uint64_t tmp;
+

target/s390x/diag.c line 44

Quoted 10 lines
target/s390x/diag.c +10 -0
+    /* TODO: Unlock with new QEMU machine. */
+    if (false) {
+        s390_program_interrupt(env, PGM_OPERATION, ra);
+        return;
+    }
+
+    /*
+     * There also seems to be subcode "0xc", which stores the size of the
+     * first chunk and the total size to r1/r2. It's only used by very old
+     * Linux, so don't implement it.

FWIW,
https://www-01.ibm.com/servers/resourcelink/svc0302a.nsf/pages/zVMV7R1sc246272/$file/hcpb4_v7r1.pdf
seems to list the available subcodes. Anything but 0xc and 0x10 is for 24/31 bit only, so we can safely ignore them. Not sure what we want to do with 0xc: it is supposed to "Return the highest addressable byte of virtual storage in the host-primary address space, including named saved systems and saved segments", so returning the end of the address space should be easy enough, but not very useful.

target/s390x/diag.c line 53

Quoted 9 lines
target/s390x/diag.c +9 -0
+     */
+    if ((r1 & 1) || subcode != 0x10) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+    addr = env->regs[r1];
+    length = env->regs[r1 + 1];
+
+    /* FIXME: Somebody with documentation should fix this. */

Doc mentioned above says for specification exception:

"For subcode X'10':
• Rx is not an even-numbered register.
• The address contained in Rx is not on a quadword boundary.
• The length contained in Rx+1 is not a positive multiple of 16."

target/s390x/diag.c line 60

Quoted 7 lines
target/s390x/diag.c +7 -0
+    if (!QEMU_IS_ALIGNED(addr, 16) || !QEMU_IS_ALIGNED(length, 16)) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */
+    if (!length) {

Probably specification exception as well?

target/s390x/diag.c line 65

Quoted 5 lines
target/s390x/diag.c +5 -0
+        setcc(cpu, 3);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */

For access exception:

"For subcode X'10', an error occurred trying to store the extent information into the guest's output area."

Quoted 20 lines
target/s390x/diag.c +17 -0
+    if (!address_space_access_valid(&address_space_memory, addr, length, true,
+                                    MEMTXATTRS_UNSPECIFIED)) {
+        s390_program_interrupt(env, PGM_ADDRESSING, ra);
+        return;
+    }
+
+    /* Indicate our initial memory ([0 .. ram_size - 1]) */
+    tmp = cpu_to_be64(0);
+    cpu_physical_memory_write(addr, &tmp, sizeof(tmp));
+    tmp = cpu_to_be64(initial_ram_size - 1);
+    cpu_physical_memory_write(addr + sizeof(tmp), &tmp, sizeof(tmp));
+
+    /* Exactly one entry was stored. */
+    env->regs[r3] = 1;
+    setcc(cpu, 0);
+}
+
 int handle_diag_288(CPUS390XState *env, uint64_t r1, uint64_t r3)
 {
     uint64_t func = env->regs[r1];

(...)

Quoted 7 lines
target/s390x/misc_helper.c +0 -0
index 58dbc023eb..d7274eb320 100644
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

Quoted 15 lines
target/s390x/translate.c +4 -0
index 4f6f1e31cd..6bb8b6e513 100644
@@ -2398,6 +2398,10 @@ static DisasJumpType op_diag(DisasContext *s, DisasOps *o)
     TCGv_i32 func_code = tcg_const_i32(get_field(s, i2));
 
     gen_helper_diag(cpu_env, r1, r3, func_code);
+    /* Only some diags modify the CC. */
+    if (get_field(s, i2) == 0x260) {
+        set_cc_static(s);
+    }
 
     tcg_temp_free_i32(func_code);
     tcg_temp_free_i32(r3);
#4 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 60 lines · Cornelia Huck
Quoted 46 lines · On Wed, 8 Jul 2020 20:51:32 +0200 David Hildenbrand <[email protected]> wrote:

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

---
 target/s390x/diag.c        | 57 ++++++++++++++++++++++++++++++++++++++
 target/s390x/internal.h    |  2 ++
 target/s390x/kvm.c         | 11 ++++++++
 target/s390x/misc_helper.c |  6 ++++
 target/s390x/translate.c   |  4 +++
 5 files changed, 80 insertions(+)
target/s390x/diag.c +9 -0
index 1a48429564..c3b1e24b2c 100644
@@ -23,6 +23,63 @@
 #include "hw/s390x/pv.h"
 #include "kvm_s390x.h"
 
+void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
+{
+    MachineState *ms = MACHINE(qdev_get_machine());
+    const ram_addr_t initial_ram_size = ms->ram_size;
+    const uint64_t subcode = env->regs[r3];
+    S390CPU *cpu = env_archcpu(env);
+    ram_addr_t addr, length;
+    uint64_t tmp;
+

target/s390x/diag.c line 44

Quoted 10 lines
target/s390x/diag.c +10 -0
+    /* TODO: Unlock with new QEMU machine. */
+    if (false) {
+        s390_program_interrupt(env, PGM_OPERATION, ra);
+        return;
+    }
+
+    /*
+     * There also seems to be subcode "0xc", which stores the size of the
+     * first chunk and the total size to r1/r2. It's only used by very old
+     * Linux, so don't implement it.
Quoted 8 lines

FWIW,
https://www-01.ibm.com/servers/resourcelink/svc0302a.nsf/pages/zVMV7R1sc246272/$file/hcpb4_v7r1.pdf
seems to list the available subcodes. Anything but 0xc and 0x10 is for 24/31 bit only, so we can safely ignore them. Not sure what we want to do with 0xc: it is supposed to "Return the highest addressable byte of virtual storage in the host-primary address space, including named saved systems and saved segments", so returning the end of the address space should be easy enough, but not very useful.

Thanks for the link to the documentation! Either my google search skills are bad or that stuff is just hard to find :) I'll have a look and see how to make sense of 0xc. Smells like "maxram_size - 1" indeed.

Quoted 26 lines

target/s390x/diag.c line 53

Quoted 10 lines
target/s390x/diag.c +9 -0
+     */
+    if ((r1 & 1) || subcode != 0x10) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+    addr = env->regs[r1];
+    length = env->regs[r1 + 1];
+
+    /* FIXME: Somebody with documentation should fix this. */

Doc mentioned above says for specification exception:

"For subcode X'10':
• Rx is not an even-numbered register.
• The address contained in Rx is not on a quadword boundary.
• The length contained in Rx+1 is not a positive multiple of 16."

target/s390x/diag.c line 60

Quoted 7 lines
target/s390x/diag.c +7 -0
+    if (!QEMU_IS_ALIGNED(addr, 16) || !QEMU_IS_ALIGNED(length, 16)) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */
+    if (!length) {
Quoted 1 line

Probably specification exception as well?

Yeah I'll add "|| !length" above.

Quoted 7 lines

target/s390x/diag.c line 65

Quoted 6 lines
target/s390x/diag.c +5 -0
+        setcc(cpu, 3);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */
Quoted 5 lines

For access exception:

"For subcode X'10', an error occurred trying to store the extent information into the guest's output area."

Okay, looks good then!

Quoted 30 lines
Quoted 20 lines
target/s390x/diag.c +17 -0
+    if (!address_space_access_valid(&address_space_memory, addr, length, true,
+                                    MEMTXATTRS_UNSPECIFIED)) {
+        s390_program_interrupt(env, PGM_ADDRESSING, ra);
+        return;
+    }
+
+    /* Indicate our initial memory ([0 .. ram_size - 1]) */
+    tmp = cpu_to_be64(0);
+    cpu_physical_memory_write(addr, &tmp, sizeof(tmp));
+    tmp = cpu_to_be64(initial_ram_size - 1);
+    cpu_physical_memory_write(addr + sizeof(tmp), &tmp, sizeof(tmp));
+
+    /* Exactly one entry was stored. */
+    env->regs[r3] = 1;
+    setcc(cpu, 0);
+}
+
 int handle_diag_288(CPUS390XState *env, uint64_t r1, uint64_t r3)
 {
     uint64_t func = env->regs[r1];

(...)

Quoted 7 lines
target/s390x/misc_helper.c +0 -0
index 58dbc023eb..d7274eb320 100644
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
Quoted 14 lines

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

Interesting. Adding in onto my todo list.

Thanks again!

Signature · 3 lines
Thanks,

David / dhildenb
#5 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

In reply to Cornelia Huck
Quoted 137 lines · Cornelia Huck
Quoted 46 lines · On Wed, 8 Jul 2020 20:51:32 +0200 David Hildenbrand <[email protected]> wrote:

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

---
 target/s390x/diag.c        | 57 ++++++++++++++++++++++++++++++++++++++
 target/s390x/internal.h    |  2 ++
 target/s390x/kvm.c         | 11 ++++++++
 target/s390x/misc_helper.c |  6 ++++
 target/s390x/translate.c   |  4 +++
 5 files changed, 80 insertions(+)
target/s390x/diag.c +9 -0
index 1a48429564..c3b1e24b2c 100644
@@ -23,6 +23,63 @@
 #include "hw/s390x/pv.h"
 #include "kvm_s390x.h"
 
+void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
+{
+    MachineState *ms = MACHINE(qdev_get_machine());
+    const ram_addr_t initial_ram_size = ms->ram_size;
+    const uint64_t subcode = env->regs[r3];
+    S390CPU *cpu = env_archcpu(env);
+    ram_addr_t addr, length;
+    uint64_t tmp;
+

target/s390x/diag.c line 44

Quoted 10 lines
target/s390x/diag.c +10 -0
+    /* TODO: Unlock with new QEMU machine. */
+    if (false) {
+        s390_program_interrupt(env, PGM_OPERATION, ra);
+        return;
+    }
+
+    /*
+     * There also seems to be subcode "0xc", which stores the size of the
+     * first chunk and the total size to r1/r2. It's only used by very old
+     * Linux, so don't implement it.

FWIW,
https://www-01.ibm.com/servers/resourcelink/svc0302a.nsf/pages/zVMV7R1sc246272/$file/hcpb4_v7r1.pdf
seems to list the available subcodes. Anything but 0xc and 0x10 is for 24/31 bit only, so we can safely ignore them. Not sure what we want to do with 0xc: it is supposed to "Return the highest addressable byte of virtual storage in the host-primary address space, including named saved systems and saved segments", so returning the end of the address space should be easy enough, but not very useful.

target/s390x/diag.c line 53

Quoted 9 lines
target/s390x/diag.c +9 -0
+     */
+    if ((r1 & 1) || subcode != 0x10) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+    addr = env->regs[r1];
+    length = env->regs[r1 + 1];
+
+    /* FIXME: Somebody with documentation should fix this. */

Doc mentioned above says for specification exception:

"For subcode X'10':
• Rx is not an even-numbered register.
• The address contained in Rx is not on a quadword boundary.
• The length contained in Rx+1 is not a positive multiple of 16."

target/s390x/diag.c line 60

Quoted 7 lines
target/s390x/diag.c +7 -0
+    if (!QEMU_IS_ALIGNED(addr, 16) || !QEMU_IS_ALIGNED(length, 16)) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */
+    if (!length) {

Probably specification exception as well?

target/s390x/diag.c line 65

Quoted 5 lines
target/s390x/diag.c +5 -0
+        setcc(cpu, 3);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */

For access exception:

"For subcode X'10', an error occurred trying to store the extent information into the guest's output area."

Quoted 20 lines
target/s390x/diag.c +17 -0
+    if (!address_space_access_valid(&address_space_memory, addr, length, true,
+                                    MEMTXATTRS_UNSPECIFIED)) {
+        s390_program_interrupt(env, PGM_ADDRESSING, ra);
+        return;
+    }
+
+    /* Indicate our initial memory ([0 .. ram_size - 1]) */
+    tmp = cpu_to_be64(0);
+    cpu_physical_memory_write(addr, &tmp, sizeof(tmp));
+    tmp = cpu_to_be64(initial_ram_size - 1);
+    cpu_physical_memory_write(addr + sizeof(tmp), &tmp, sizeof(tmp));
+
+    /* Exactly one entry was stored. */
+    env->regs[r3] = 1;
+    setcc(cpu, 0);
+}
+
 int handle_diag_288(CPUS390XState *env, uint64_t r1, uint64_t r3)
 {
     uint64_t func = env->regs[r1];

(...)

Quoted 7 lines
target/s390x/misc_helper.c +0 -0
index 58dbc023eb..d7274eb320 100644
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
Quoted 14 lines

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

That's also a BUG in kvm then?

int kvm_s390_handle_diag(struct kvm_vcpu *vcpu)
{
...

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

...
}

Signature · 3 lines
Thanks,

David / dhildenb
#6 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 155 lines · David Hildenbrand
Quoted 137 lines · Cornelia Huck
Quoted 46 lines · On Wed, 8 Jul 2020 20:51:32 +0200 David Hildenbrand <[email protected]> wrote:

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

---
 target/s390x/diag.c        | 57 ++++++++++++++++++++++++++++++++++++++
 target/s390x/internal.h    |  2 ++
 target/s390x/kvm.c         | 11 ++++++++
 target/s390x/misc_helper.c |  6 ++++
 target/s390x/translate.c   |  4 +++
 5 files changed, 80 insertions(+)
target/s390x/diag.c +9 -0
index 1a48429564..c3b1e24b2c 100644
@@ -23,6 +23,63 @@
 #include "hw/s390x/pv.h"
 #include "kvm_s390x.h"
 
+void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
+{
+    MachineState *ms = MACHINE(qdev_get_machine());
+    const ram_addr_t initial_ram_size = ms->ram_size;
+    const uint64_t subcode = env->regs[r3];
+    S390CPU *cpu = env_archcpu(env);
+    ram_addr_t addr, length;
+    uint64_t tmp;
+

target/s390x/diag.c line 44

Quoted 10 lines
target/s390x/diag.c +10 -0
+    /* TODO: Unlock with new QEMU machine. */
+    if (false) {
+        s390_program_interrupt(env, PGM_OPERATION, ra);
+        return;
+    }
+
+    /*
+     * There also seems to be subcode "0xc", which stores the size of the
+     * first chunk and the total size to r1/r2. It's only used by very old
+     * Linux, so don't implement it.

FWIW,
https://www-01.ibm.com/servers/resourcelink/svc0302a.nsf/pages/zVMV7R1sc246272/$file/hcpb4_v7r1.pdf
seems to list the available subcodes. Anything but 0xc and 0x10 is for 24/31 bit only, so we can safely ignore them. Not sure what we want to do with 0xc: it is supposed to "Return the highest addressable byte of virtual storage in the host-primary address space, including named saved systems and saved segments", so returning the end of the address space should be easy enough, but not very useful.

target/s390x/diag.c line 53

Quoted 9 lines
target/s390x/diag.c +9 -0
+     */
+    if ((r1 & 1) || subcode != 0x10) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+    addr = env->regs[r1];
+    length = env->regs[r1 + 1];
+
+    /* FIXME: Somebody with documentation should fix this. */

Doc mentioned above says for specification exception:

"For subcode X'10':
• Rx is not an even-numbered register.
• The address contained in Rx is not on a quadword boundary.
• The length contained in Rx+1 is not a positive multiple of 16."

target/s390x/diag.c line 60

Quoted 7 lines
target/s390x/diag.c +7 -0
+    if (!QEMU_IS_ALIGNED(addr, 16) || !QEMU_IS_ALIGNED(length, 16)) {
+        s390_program_interrupt(env, PGM_SPECIFICATION, ra);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */
+    if (!length) {

Probably specification exception as well?

target/s390x/diag.c line 65

Quoted 5 lines
target/s390x/diag.c +5 -0
+        setcc(cpu, 3);
+        return;
+    }
+
+    /* FIXME: Somebody with documentation should fix this. */

For access exception:

"For subcode X'10', an error occurred trying to store the extent information into the guest's output area."

Quoted 20 lines
target/s390x/diag.c +17 -0
+    if (!address_space_access_valid(&address_space_memory, addr, length, true,
+                                    MEMTXATTRS_UNSPECIFIED)) {
+        s390_program_interrupt(env, PGM_ADDRESSING, ra);
+        return;
+    }
+
+    /* Indicate our initial memory ([0 .. ram_size - 1]) */
+    tmp = cpu_to_be64(0);
+    cpu_physical_memory_write(addr, &tmp, sizeof(tmp));
+    tmp = cpu_to_be64(initial_ram_size - 1);
+    cpu_physical_memory_write(addr + sizeof(tmp), &tmp, sizeof(tmp));
+
+    /* Exactly one entry was stored. */
+    env->regs[r3] = 1;
+    setcc(cpu, 0);
+}
+
 int handle_diag_288(CPUS390XState *env, uint64_t r1, uint64_t r3)
 {
     uint64_t func = env->regs[r1];

(...)

Quoted 7 lines
target/s390x/misc_helper.c +0 -0
index 58dbc023eb..d7274eb320 100644
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
Quoted 14 lines

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

That's also a BUG in kvm then?

Quoted 8 lines

int kvm_s390_handle_diag(struct kvm_vcpu *vcpu)
{
...

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

...
}

But OTOH, it does not sound sane if user space can bypass the OS to yield the CPU ... so this might just be a wrong documentation. All DIAGs should be privileged IIRC.

Signature · 3 lines
Thanks,

David / dhildenb
#7 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 38 lines · On Fri, 10 Jul 2020 10:41:33 +0200 David Hildenbrand <[email protected]> wrote:
Quoted 28 lines · David Hildenbrand
Quoted 10 lines · Cornelia Huck
Quoted 7 lines · On Wed, 8 Jul 2020 20:51:32 +0200 David Hildenbrand <[email protected]> wrote:
target/s390x/misc_helper.c +0 -0
index 58dbc023eb..d7274eb320 100644
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
Quoted 14 lines

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();  

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

That's also a BUG in kvm then?

Quoted 8 lines

int kvm_s390_handle_diag(struct kvm_vcpu *vcpu)
{
...

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

...
}

Quoted 3 lines

But OTOH, it does not sound sane if user space can bypass the OS to yield the CPU ... so this might just be a wrong documentation. All DIAGs should be privileged IIRC.

Maybe not all of them, but the diag 0x44 case is indeed odd. No idea what is documented for its use on LPAR (I don't think that document is public.)

#8 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

In reply to David Hildenbrand
Quoted 22 lines · David Hildenbrand
Quoted 5 lines
Quoted 5 lines
target/s390x/misc_helper.c +0 -0
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
Quoted 14 lines

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

That's also a BUG in kvm then?

Quoted 7 lines

int kvm_s390_handle_diag(struct kvm_vcpu *vcpu)
{
...

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

...
}

diag 44 gives a PRIVOP on LPAR, so I think this is fine.

#9 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 32 lines · On Mon, 13 Jul 2020 13:54:41 +0200 Christian Borntraeger…
Quoted 22 lines · David Hildenbrand
Quoted 5 lines
Quoted 5 lines
target/s390x/misc_helper.c +0 -0
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
Quoted 14 lines

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();  

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

That's also a BUG in kvm then?

Quoted 7 lines

int kvm_s390_handle_diag(struct kvm_vcpu *vcpu)
{
...

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

...
}

Quoted 2 lines

diag 44 gives a PRIVOP on LPAR, so I think this is fine.

Seems like a bug/inconsistency in CP (or its documentation), then.

#10 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 35 lines · Cornelia Huck
Quoted 32 lines · On Mon, 13 Jul 2020 13:54:41 +0200 Christian Borntraeger…
Quoted 29 lines · David Hildenbrand
Quoted 5 lines
Quoted 5 lines
target/s390x/misc_helper.c +0 -0
@@ -116,6 +116,12 @@ void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num)
     uint64_t r;
 
Quoted 14 lines

target/s390x/misc_helper.c void HELPER(diag)(CPUS390XState *env, uint32_t r1, uint32_t r3, uint32_t num) · line 127

Quoted 10 lines
target/s390x/misc_helper.c +6 -0
     switch (num) {
+    case 0x260:
+        qemu_mutex_lock_iothread();
+        handle_diag_260(env, r1, r3, GETPC());
+        qemu_mutex_unlock_iothread();
+        r = 0;
+        break;
     case 0x500:
         /* KVM hypercall */
         qemu_mutex_lock_iothread();  

Looking at the doc referenced above, it seems that we treat every diag call as privileged under tcg; but it seems that 0x44 isn't? (Unrelated to your patch; maybe I'm misreading.)

That's also a BUG in kvm then?

int kvm_s390_handle_diag(struct kvm_vcpu *vcpu)
{
...

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

...
}

Quoted 4 lines
Quoted 2 lines

diag 44 gives a PRIVOP on LPAR, so I think this is fine.

Seems like a bug/inconsistency in CP (or its documentation), then.

Yes.

.globl main
main:

        diag 0,0,0x44
        svc 1

also crashes under z/VM with an illegal op.

#11 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

In reply to David Hildenbrand
Quoted 13 lines · David Hildenbrand

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

Quoted 8 lines

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

I have not looked into this, so this is purely a question.

Is there a way to hotplug virtio-mem memory beyond the initial size of the memory as specified by the initial sclp)? then we could avoid doing this platform specfic diag260?
the only issue I see is when we need to go beyond 4TB due to the page table upgrade in the kernel.

FWIW diag 260 is publicly documented.

#12 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 22 lines · Christian Borntraeger
Quoted 20 lines · David Hildenbrand

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

Quoted 7 lines
Quoted 1 line

I have not looked into this, so this is purely a question.

Is there a way to hotplug virtio-mem memory beyond the initial size of the memory as specified by the initial sclp)? then we could avoid doing this platform specfic diag260?

We need a way to tell the guest about the maximum possible PFN, so it can prepare for that. E.g. on x86-64 this is usually done via ACPI SRAT tables. On s390x, the only way I see is using a combination of diag260, without introducing any other new mechanisms.

Currently Linux selects 3. vs 4 level page tables based on that size (I think that's what you were referring to with the 4TB limit). I can see that kasan also does some magic based on the value ("populate kasan shadow for untracked memory"), but did not look into the details. I *think* kasan will never be able to track that memory, but am not completely sure.

I'd like to avoid something as you propose (that's why I searched and discovered diag260 after all :) ), especially to not silently break in the future, when other assumptions based on that value are introduced.

E.g., on my z/VM LinuxOne Community Cloud machine, diag260 gets used as default, so it does not seem to be a corner case mechanism nowadays.

Quoted 4 lines

the only issue I see is when we need to go beyond 4TB due to the page table upgrade in the kernel.

FWIW diag 260 is publicly documented.

Yeah, Conny pointed me at the doc - makes things easier :)

Signature · 3 lines
Thanks,

David / dhildenb
#13 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 43 lines · David Hildenbrand
Quoted 22 lines · Christian Borntraeger
Quoted 20 lines · David Hildenbrand

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

Quoted 7 lines
Quoted 1 line

I have not looked into this, so this is purely a question.

Is there a way to hotplug virtio-mem memory beyond the initial size of the memory as specified by the initial sclp)? then we could avoid doing this platform specfic diag260?

We need a way to tell the guest about the maximum possible PFN, so it can prepare for that. E.g. on x86-64 this is usually done via ACPI SRAT tables. On s390x, the only way I see is using a combination of diag260, without introducing any other new mechanisms.

Currently Linux selects 3. vs 4 level page tables based on that size (I think that's what you were referring to with the 4TB limit). I can see that kasan also does some magic based on the value ("populate kasan shadow for untracked memory"), but did not look into the details. I *think* kasan will never be able to track that memory, but am not completely sure.

Quoted 7 lines

I'd like to avoid something as you propose (that's why I searched and discovered diag260 after all :) ), especially to not silently break in the future, when other assumptions based on that value are introduced.

E.g., on my z/VM LinuxOne Community Cloud machine, diag260 gets used as default, so it does not seem to be a corner case mechanism nowadays.

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Signature · 3 lines
Thanks,

David / dhildenb
#14 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 48 lines · David Hildenbrand
Quoted 47 lines · David Hildenbrand
Quoted 22 lines · Christian Borntraeger
Quoted 20 lines · David Hildenbrand

Let's implement the "storage configuration" part of diag260. This diag is found under z/VM, to indicate usable chunks of memory tot he guest OS. As I don't have access to documentation, I have no clue what the actual error cases are, and which other stuff we could eventually query using this interface. Somebody with access to documentation should fix this. This implementation seems to work with Linux guests just fine.

The Linux kernel supports diag260 to query the available memory since v4.20. Older kernels / kvm-unit-tests will later fail to run in such a VM (with maxmem being defined and bigger than the memory size, e.g., "-m
2G,maxmem=4G"), just as if support for SCLP storage information is not
implemented. They will fail to detect the actual initial memory size.

This interface allows us to expose the maximum ramsize via sclp and the initial ramsize via diag260 - without having to mess with the memory increment size and having to align the initial memory size to it.

This is a preparation for memory device support. We'll unlock the implementation with a new QEMU machine that supports memory devices.

Quoted 7 lines
Quoted 1 line

I have not looked into this, so this is purely a question.

Is there a way to hotplug virtio-mem memory beyond the initial size of the memory as specified by the initial sclp)? then we could avoid doing this platform specfic diag260?

We need a way to tell the guest about the maximum possible PFN, so it can prepare for that. E.g. on x86-64 this is usually done via ACPI SRAT tables. On s390x, the only way I see is using a combination of diag260, without introducing any other new mechanisms.

Currently Linux selects 3. vs 4 level page tables based on that size (I think that's what you were referring to with the 4TB limit). I can see that kasan also does some magic based on the value ("populate kasan shadow for untracked memory"), but did not look into the details. I *think* kasan will never be able to track that memory, but am not completely sure.

I'd like to avoid something as you propose (that's why I searched and discovered diag260 after all :) ), especially to not silently break in the future, when other assumptions based on that value are introduced.

Quoted 8 lines
Quoted 3 lines

E.g., on my z/VM LinuxOne Community Cloud machine, diag260 gets used as default, so it does not seem to be a corner case mechanism nowadays.

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Implemented it, seems to work fine.

Signature · 3 lines
Thanks,

David / dhildenb
#15 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 6 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.
Also: why do you want to change this?

#16 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 10 lines · Heiko Carstens
Quoted 6 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.

Which value exactly are you referencing? diag 0xc returns two values. One of them seems to do exactly what we need.

See
https://github.com/davidhildenbrand/linux/commit/a235f9fb20df7c04ae89bc0d134332d1a01842c7

for my current Linux approach.

Quoted 1 line

Also: why do you want to change this

Which change exactly do you mean?

If we limit the value returned via SCLP to initial memory, we cannot break any guest (e.g., Linux pre 4.2, kvm-unit-tests). diag260 is then purely optional.

Signature · 3 lines
Thanks,

David / dhildenb
#17 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 7 lines · David Hildenbrand
Quoted 6 lines · Heiko Carstens
Quoted 5 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Quoted 7 lines
Quoted 4 lines
Quoted 1 line

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.

Which value exactly are you referencing? diag 0xc returns two values. One of them seems to do exactly what we need.

Maybe I'm missing something as usual, but to me this
--------
Usage Notes:
...
2. If the RESERVED or STANDBY option was used on the DEFINE STORAGE command to configure reserved or standby storage for a guest, the values returned in Rx and Ry will be the current values, but these values can change dynamically depending on the options specified and any dynamic storage reconfiguration (DSR) changes initiated by the guest.
--------
reads like it is not doing what you want. That is: it does *not* include standby memory and therefore will not return the highest possible pfn.

#18 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 20 lines

Am 10.07.2020 um 17:43 schrieb Heiko Carstens <[email protected]>:

Quoted 7 lines · David Hildenbrand
Quoted 7 lines · Heiko Carstens
Quoted 5 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Quoted 7 lines
Quoted 4 lines
Quoted 1 line

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.

Which value exactly are you referencing? diag 0xc returns two values. One of them seems to do exactly what we need.

Quoted 15 lines

Maybe I'm missing something as usual, but to me this
--------
Usage Notes:
...
2. If the RESERVED or STANDBY option was used on the DEFINE STORAGE command to configure reserved or standby storage for a guest, the values returned in Rx and Ry will be the current values, but these values can change dynamically depending on the options specified and any dynamic storage reconfiguration (DSR) changes initiated by the guest.
--------
reads like it is not doing what you want. That is: it does *not* include standby memory and therefore will not return the highest possible pfn.

Ah, yes. See the kernel patch, I take the max of both values (SCLP, diag260(0xc)) values.

Anyhow, what would be your recommendation?

Thanks!

#19 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

In reply to David Hildenbrand
Quoted 20 lines · David Hildenbrand
Quoted 10 lines · Heiko Carstens
Quoted 6 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.

Which value exactly are you referencing? diag 0xc returns two values. One of them seems to do exactly what we need.

See
https://github.com/davidhildenbrand/linux/commit/a235f9fb20df7c04ae89bc0d134332d1a01842c7

for my current Linux approach.

Quoted 7 lines
Quoted 1 line

Also: why do you want to change this

Which change exactly do you mean?

If we limit the value returned via SCLP to initial memory, we cannot break any guest (e.g., Linux pre 4.2, kvm-unit-tests). diag260 is then purely optional.

Ok, now I see the context. Christian added my just to cc on this specific patch.
So if I understand you correctly, then you want to use diag 260 in order to figure out how much memory is _potentially_ available for a guest?

This does not fit to the current semantics, since diag 260 returns the address of the highest *currently* accessible address. That is: it does explicitly *not* include standby memory or anything else that might potentially be there.

So you would need a different interface to tell the guest about your new hotplug memory interface. If sclp does not work, then maybe a new diagnose(?).

#20 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 27 lines

Am 13.07.2020 um 11:12 schrieb Heiko Carstens <[email protected]>:

Quoted 22 lines · David Hildenbrand
Quoted 11 lines · Heiko Carstens
Quoted 6 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.

Which value exactly are you referencing? diag 0xc returns two values. One of them seems to do exactly what we need.

See
https://github.com/davidhildenbrand/linux/commit/a235f9fb20df7c04ae89bc0d134332d1a01842c7

for my current Linux approach.

Quoted 1 line

Also: why do you want to change this

Quoted 8 lines
Quoted 5 lines

Which change exactly do you mean?

If we limit the value returned via SCLP to initial memory, we cannot break any guest (e.g., Linux pre 4.2, kvm-unit-tests). diag260 is then purely optional.

Ok, now I see the context. Christian added my just to cc on this specific patch.

I tried to Cc you an all patches but the mail bounced with unknown address (maybe I messed up).

Quoted 3 lines

So if I understand you correctly, then you want to use diag 260 in order to figure out how much memory is _potentially_ available for a guest?

Yes, exactly.

Quoted 5 lines

This does not fit to the current semantics, since diag 260 returns the address of the highest *currently* accessible address. That is: it does explicitly *not* include standby memory or anything else that might potentially be there.

The confusing part is that it talks about „adressible“ and not „accessible“. Now that I understood the „DEFINE STORAGE ...“ example, it makes sense that the values change with reserved/standby memory.

I agree that reusing that interface might not be what we want. I just seemed too easy to avoid creating something new :)

Quoted 5 lines

So you would need a different interface to tell the guest about your new hotplug memory interface. If sclp does not work, then maybe a new diagnose(?).

Yes, I think a new Diagnose makes sense. I‘ll have a look next week to figure out which codes/subcodes we could use. @Christian @Conny any ideas/pointers?

#21 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 55 lines · David Hildenbrand
Quoted 27 lines

Am 13.07.2020 um 11:12 schrieb Heiko Carstens <[email protected]>:

Quoted 22 lines · David Hildenbrand
Quoted 11 lines · Heiko Carstens
Quoted 6 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.

Which value exactly are you referencing? diag 0xc returns two values. One of them seems to do exactly what we need.

See
https://github.com/davidhildenbrand/linux/commit/a235f9fb20df7c04ae89bc0d134332d1a01842c7

for my current Linux approach.

Quoted 1 line

Also: why do you want to change this

Quoted 8 lines
Quoted 5 lines

Which change exactly do you mean?

If we limit the value returned via SCLP to initial memory, we cannot break any guest (e.g., Linux pre 4.2, kvm-unit-tests). diag260 is then purely optional.

Ok, now I see the context. Christian added my just to cc on this specific patch.

I tried to Cc you an all patches but the mail bounced with unknown address (maybe I messed up).

Quoted 3 lines

So if I understand you correctly, then you want to use diag 260 in order to figure out how much memory is _potentially_ available for a guest?

Yes, exactly.

Quoted 5 lines

This does not fit to the current semantics, since diag 260 returns the address of the highest *currently* accessible address. That is: it does explicitly *not* include standby memory or anything else that might potentially be there.

The confusing part is that it talks about „adressible“ and not „accessible“. Now that I understood the „DEFINE STORAGE ...“ example, it makes sense that the values change with reserved/standby memory.

I agree that reusing that interface might not be what we want. I just seemed too easy to avoid creating something new :)

Quoted 1 line
Quoted 6 lines
Quoted 4 lines

So you would need a different interface to tell the guest about your new hotplug memory interface. If sclp does not work, then maybe a new diagnose(?).

Yes, I think a new Diagnose makes sense. I‘ll have a look next week to figure out which codes/subcodes we could use. @Christian @Conny any ideas/pointers?>

Wouldnt sclp be the right thing to provide the max increment number? (and thus the max memory address) And then (when I got the discussion right) use diag 260 to get the _current_ value.

#22 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 61 lines · Christian Borntraeger
Quoted 60 lines · David Hildenbrand
Quoted 27 lines

Am 13.07.2020 um 11:12 schrieb Heiko Carstens <[email protected]>:

Quoted 22 lines · David Hildenbrand
Quoted 11 lines · Heiko Carstens
Quoted 6 lines · David Hildenbrand
Quoted 4 lines

Note: Reading about diag260 subcode 0xc, we could modify Linux to query the maximum possible pfn via diag260 0xc. Then, we maybe could avoid indicating maxram size via SCLP, and keep diag260-unaware OSs keep working as before. Thoughts?

Implemented it, seems to work fine.

The returned value would not include standby/reserved memory within z/VM. So this seems not to work.

Which value exactly are you referencing? diag 0xc returns two values. One of them seems to do exactly what we need.

See
https://github.com/davidhildenbrand/linux/commit/a235f9fb20df7c04ae89bc0d134332d1a01842c7

for my current Linux approach.

Quoted 1 line

Also: why do you want to change this

Quoted 8 lines
Quoted 5 lines

Which change exactly do you mean?

If we limit the value returned via SCLP to initial memory, we cannot break any guest (e.g., Linux pre 4.2, kvm-unit-tests). diag260 is then purely optional.

Ok, now I see the context. Christian added my just to cc on this specific patch.

I tried to Cc you an all patches but the mail bounced with unknown address (maybe I messed up).

Quoted 3 lines

So if I understand you correctly, then you want to use diag 260 in order to figure out how much memory is _potentially_ available for a guest?

Yes, exactly.

Quoted 5 lines

This does not fit to the current semantics, since diag 260 returns the address of the highest *currently* accessible address. That is: it does explicitly *not* include standby memory or anything else that might potentially be there.

The confusing part is that it talks about „adressible“ and not „accessible“. Now that I understood the „DEFINE STORAGE ...“ example, it makes sense that the values change with reserved/standby memory.

I agree that reusing that interface might not be what we want. I just seemed too easy to avoid creating something new :)

Quoted 5 lines

So you would need a different interface to tell the guest about your new hotplug memory interface. If sclp does not work, then maybe a new diagnose(?).

Quoted 4 lines
Quoted 1 line

Yes, I think a new Diagnose makes sense. I‘ll have a look next week to figure out which codes/subcodes we could use. @Christian @Conny any ideas/pointers?>

Wouldnt sclp be the right thing to provide the max increment number? (and thus the max memory address) And then (when I got the discussion right) use diag 260 to get the _current_ value.

So, in summary, we want to indicate to the guest a memory region that will be used to place memory devices ("device memory region"). The region might have holes and the memory within this region might have different semantics than ordinary system memory. Memory that belongs to memory devices should only be detected+used if the guest OS has support for them (e.g., virtio-mem, virtio-pmem, ...). An unmodified guest (e.g., no virtio-mem driver) should not accidentally make use of such memory.

We need a way to
a) Tell the guest about boot memory (currently ram_size)
b) Tell the guest about the maximum possible ram address, including device memory. (We could also indicate the special "device memory region" explicitly)

AFAIK, we have three options:

1. Indicate maxram_size via SCLP, indicate ram_size via diag260(0x10)

This is what this series (RFCv1 does).

Advantages:
- No need for a new diag. No need for memory sensing kernel changes. Disadvantages
- Older guests without support for diag260 (<v4.2, kvm-unit-tests) will assume all memory is accessible. Bad.
- The semantics of the value returned in ry via diag260(0xc) is somewhat unclear. Should we return the end address of the highest memory device? OTOH, an unmodified guest OS (without support for memory devices) should not have to care at all about any such memory.
- If we ever want to also support standby memory, we might be in trouble. (see below)

2. Indicate ram_size via SCLP, indicate device memory region (currently maxram_size) via new DIAG

Advantages:
- Unmodified guests won't use/sense memory belonging to memory devices.
- We can later have standby memory + memory devices co-exist. Disadvantages
- Need a new DIAG.

3. Indicate maxram_size and ram_size via SCLP (using the SCLP standby memory)

I did not look into the details, because -ENODOCUMENTATION. At least we would run into some alignment issues (again, having to align ram_size/maxram_size to storage increments - which would no longer be 1MB). We would run into issues later, trying to also support standby memory.

I guess 1) would mostly work, one just has to run a suitable guest inside the VM. This is no different to running under z/VM where querying diag260 is required. The nice thing about 2) would be, that we can easily implement standby memory. Something like:

-m 2G,maxram_size=20G,standbyram_size=4G

[ 2G boot RAM ][ 4G standby RAM ][ 14G device memory ]
^ via SCLP maximum increment
^ via new DIAG

Signature · 3 lines
Thanks,

David / dhildenb
#23 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 19 lines · David Hildenbrand

So, in summary, we want to indicate to the guest a memory region that will be used to place memory devices ("device memory region"). The region might have holes and the memory within this region might have different semantics than ordinary system memory. Memory that belongs to memory devices should only be detected+used if the guest OS has support for them (e.g., virtio-mem, virtio-pmem, ...). An unmodified guest (e.g., no virtio-mem driver) should not accidentally make use of such memory.

We need a way to
a) Tell the guest about boot memory (currently ram_size)
b) Tell the guest about the maximum possible ram address, including device memory. (We could also indicate the special "device memory region" explicitly)

AFAIK, we have three options:

1. Indicate maxram_size via SCLP, indicate ram_size via diag260(0x10)

Quoted 7 lines

This is what this series (RFCv1 does).

Advantages:
- No need for a new diag. No need for memory sensing kernel changes. Disadvantages
- Older guests without support for diag260 (<v4.2, kvm-unit-tests) will assume all memory is accessible. Bad.

Why would old guests assume that?

At least in v4.1 the kernel will calculate the max address by using increment size * increment number and then test if *each* increment is available with tprot.

Quoted 4 lines

- The semantics of the value returned in ry via diag260(0xc) is somewhat unclear. Should we return the end address of the highest memory device? OTOH, an unmodified guest OS (without support for memory devices) should not have to care at all about any such memory.

I'm confused. The kernel currently only uses diag260(0x10). How is diag260(0xc) relevant here?

Quoted 7 lines

3. Indicate maxram_size and ram_size via SCLP (using the SCLP standby memory)

I did not look into the details, because -ENODOCUMENTATION. At least we would run into some alignment issues (again, having to align ram_size/maxram_size to storage increments - which would no longer be 1MB). We would run into issues later, trying to also support standby memory.

That doesn't make sense to me: either support memory hotplug via sclp/standby memory, or with your new method. But trying to support both.. what's the use case?

#24 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 28 lines · Heiko Carstens
Quoted 19 lines · David Hildenbrand

So, in summary, we want to indicate to the guest a memory region that will be used to place memory devices ("device memory region"). The region might have holes and the memory within this region might have different semantics than ordinary system memory. Memory that belongs to memory devices should only be detected+used if the guest OS has support for them (e.g., virtio-mem, virtio-pmem, ...). An unmodified guest (e.g., no virtio-mem driver) should not accidentally make use of such memory.

We need a way to
a) Tell the guest about boot memory (currently ram_size)
b) Tell the guest about the maximum possible ram address, including device memory. (We could also indicate the special "device memory region" explicitly)

AFAIK, we have three options:

1. Indicate maxram_size via SCLP, indicate ram_size via diag260(0x10)

Quoted 7 lines

This is what this series (RFCv1 does).

Advantages:
- No need for a new diag. No need for memory sensing kernel changes. Disadvantages
- Older guests without support for diag260 (<v4.2, kvm-unit-tests) will assume all memory is accessible. Bad.

Quoted 5 lines

Why would old guests assume that?

At least in v4.1 the kernel will calculate the max address by using increment size * increment number and then test if *each* increment is available with tprot.

Yes, we do the same in kvm-unit-tests. But it's not sufficient for memory devices.

Just because a tprot succeed (for memory belonging to a memory device) does not mean the kernel should silently start to use that memory.

Note: memory devices are not just DIMMs that can be mapped to storage increments. The memory might have completely different semantics, that's why they are glued to a managing virtio device.

For example: a tprot might succeed on a memory region provided by virtio-mem, this does, however, not mean that the memory can (and should) be used by the guest.

Quoted 8 lines
Quoted 5 lines

- The semantics of the value returned in ry via diag260(0xc) is somewhat unclear. Should we return the end address of the highest memory device? OTOH, an unmodified guest OS (without support for memory devices) should not have to care at all about any such memory.

I'm confused. The kernel currently only uses diag260(0x10). How is diag260(0xc) relevant here?

We have to implement diag260(0x10) if we implement diag260(0xc), no? Or can we simply throw a specification exception?

Quoted 4 lines
Quoted 4 lines

3. Indicate maxram_size and ram_size via SCLP (using the SCLP standby memory)

Quoted 8 lines
Quoted 4 lines

I did not look into the details, because -ENODOCUMENTATION. At least we would run into some alignment issues (again, having to align ram_size/maxram_size to storage increments - which would no longer be 1MB). We would run into issues later, trying to also support standby memory.

That doesn't make sense to me: either support memory hotplug via sclp/standby memory, or with your new method. But trying to support both.. what's the use case?

Not sure if there is any, it just feels cleaner to me to separate the architectured (sclp memory/reserved/standby) bits that specify a semantic when used via rnmax+tprot from QEMU specific memory ranges that have special semantics.

virtio-mem is only one type of a virtio-based memory device. In the future we might want to have virtio-pmem, but there might be more ...

Signature · 3 lines
Thanks,

David / dhildenb
#25 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 10 lines · David Hildenbrand
Quoted 3 lines

At least in v4.1 the kernel will calculate the max address by using increment size * increment number and then test if *each* increment is available with tprot.

Yes, we do the same in kvm-unit-tests. But it's not sufficient for memory devices.

Just because a tprot succeed (for memory belonging to a memory device) does not mean the kernel should silently start to use that memory.

Quoted 7 lines

Note: memory devices are not just DIMMs that can be mapped to storage increments. The memory might have completely different semantics, that's why they are glued to a managing virtio device.

For example: a tprot might succeed on a memory region provided by virtio-mem, this does, however, not mean that the memory can (and should) be used by the guest.

So, are you saying that even at IPL time there might already be memory devices attached to the system? And the kernel should _not_ treat them as normal memory?

#26 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 15 lines · Heiko Carstens
Quoted 14 lines · David Hildenbrand
Quoted 3 lines

At least in v4.1 the kernel will calculate the max address by using increment size * increment number and then test if *each* increment is available with tprot.

Yes, we do the same in kvm-unit-tests. But it's not sufficient for memory devices.

Just because a tprot succeed (for memory belonging to a memory device) does not mean the kernel should silently start to use that memory.

Note: memory devices are not just DIMMs that can be mapped to storage increments. The memory might have completely different semantics, that's why they are glued to a managing virtio device.

Quoted 7 lines
Quoted 3 lines

For example: a tprot might succeed on a memory region provided by virtio-mem, this does, however, not mean that the memory can (and should) be used by the guest.

So, are you saying that even at IPL time there might already be memory devices attached to the system? And the kernel should _not_ treat them as normal memory?

Sorry if that was unclear. Yes, we can have such devices (including memory areas) on a cold boot/reboot/kexec. In addition, they might pop up at runtime (e.g., hotplugging a virtio-mem device). The device is in charge of exposing that area and deciding what to do with it.

The kernel should never treat them as normal memory (IOW, system RAM). Not during a cold boot, not during a reboot. The device driver is responsible for deciding how to use that memory (e.g., add it as system RAM), and which parts of that memory are actually valid to be used (even if a tprot might succeed it might not be valid to use just yet - I guess somewhat similar to doing a tport on a dcss area - AFAIK, you also don't want to use it like normal memory).

E.g., on x86-64, memory exposed via virtio-mem or virtio-pmem is never exposed via the e820 map. The only trace that there might be *something* now/in the future is indicated via ACPI SRAT tables. This takes currently care of indicating the maximum possible PFN.

Signature · 3 lines
Thanks,

David / dhildenb
#27 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 17 lines · David Hildenbrand
Quoted 3 lines

So, are you saying that even at IPL time there might already be memory devices attached to the system? And the kernel should _not_ treat them as normal memory?

Sorry if that was unclear. Yes, we can have such devices (including memory areas) on a cold boot/reboot/kexec. In addition, they might pop up at runtime (e.g., hotplugging a virtio-mem device). The device is in charge of exposing that area and deciding what to do with it.

The kernel should never treat them as normal memory (IOW, system RAM). Not during a cold boot, not during a reboot. The device driver is responsible for deciding how to use that memory (e.g., add it as system RAM), and which parts of that memory are actually valid to be used (even if a tprot might succeed it might not be valid to use just yet - I guess somewhat similar to doing a tport on a dcss area - AFAIK, you also don't want to use it like normal memory).

Quoted 4 lines

E.g., on x86-64, memory exposed via virtio-mem or virtio-pmem is never exposed via the e820 map. The only trace that there might be *something* now/in the future is indicated via ACPI SRAT tables. This takes currently care of indicating the maximum possible PFN.

Ok, but all of this needa to be documented somewhere. This raises a couple of questions to me:

What happens on

- IPL Clear with this special memory? Will it be detached/away afterwards?
- IPL Normal? "Obviously" it must stay otherwise kdump would never see that memory.

And when you write it's up to the device driver what to with that
memory: is there any documentation available what all of this is good for? I would assume _most likely_ this extra memory is going to be added to ZONE_MOVABLE _somehow_ so that it can be taken away also. But since it is not normal memory, like you say, I'm wondering how that is supposed to work.

As far as I can tell there would be a lot of inconsistencies in userspace interfaces which provide memory / zone information. Or I'm not getting the point of all of this at all.

So please provide more information, or a pointer to documentation.

#28 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 18 lines · Heiko Carstens
Quoted 17 lines · David Hildenbrand
Quoted 3 lines

So, are you saying that even at IPL time there might already be memory devices attached to the system? And the kernel should _not_ treat them as normal memory?

Sorry if that was unclear. Yes, we can have such devices (including memory areas) on a cold boot/reboot/kexec. In addition, they might pop up at runtime (e.g., hotplugging a virtio-mem device). The device is in charge of exposing that area and deciding what to do with it.

The kernel should never treat them as normal memory (IOW, system RAM). Not during a cold boot, not during a reboot. The device driver is responsible for deciding how to use that memory (e.g., add it as system RAM), and which parts of that memory are actually valid to be used (even if a tprot might succeed it might not be valid to use just yet - I guess somewhat similar to doing a tport on a dcss area - AFAIK, you also don't want to use it like normal memory).

Quoted 7 lines
Quoted 4 lines

E.g., on x86-64, memory exposed via virtio-mem or virtio-pmem is never exposed via the e820 map. The only trace that there might be *something* now/in the future is indicated via ACPI SRAT tables. This takes currently care of indicating the maximum possible PFN.

Ok, but all of this needa to be documented somewhere. This raises a couple of questions to me:

I assume this mostly targets virtio-mem, because the semantics of virtio-mem provided memory are extra-weird (in contrast to rather static virtio-pmem, which is essentially just an emulated NVDIMM - a disk mapped into physical memory).

Regarding documentation (some linked in the cover letter), so far I have (generic/x86-64)

1. https://virtio-mem.gitlab.io/
2. virtio spec proposal [1]
3. QEMU 910b25766b33 ("virtio-mem: Paravirtualized memory hot(un)plug")
4. Linux 5f1f79bbc9 ("virtio-mem: Paravirtualized memory hotplug")
5. Linux cover letter [2]
6. KVM forum talk [3] [4]

As your questions go quite into technical detail, and I don't feel like rewriting the doc here :) , I suggest looking at [2], 1, and 5.

Quoted 2 lines

What happens on

I'll stick to virtio-mem when answering regarding "special memory". As I noted, there might be more in the future.

Quoted 2 lines

- IPL Clear with this special memory? Will it be detached/away afterwards?

A diag308(0x3) - load clear - will usually* zap all virtio-mem provided memory (discard backing storage in the hypervisor) and logically turn the state of all virtio-mem memory inside the device-assigned memory region to "unplugged" - just as during a cold boot. The semantics of "unplugged" blocks depend on the "usable region" (see the virtio-spec if you're curious - the memory might still be accessible). Starting "fresh" with all memory logically unplugged is part of the way virtio-mem works.

* there are corner cases while a VM is getting migrated, where we cannot perform this (similar, to us not being able to clear ordinary memory during a load clear in QEMU while migrating). In this case, the memory is left untouched.

Quoted 2 lines

- IPL Normal? "Obviously" it must stay otherwise kdump would never see that memory.

Only diag308(0x3) will mess with virtio-mem memory. For the other types of resets, its left untouched. So yes, "obviously" is correct :)

Quoted 7 lines

And when you write it's up to the device driver what to with that
memory: is there any documentation available what all of this is good for? I would assume _most likely_ this extra memory is going to be added to ZONE_MOVABLE _somehow_ so that it can be taken away also. But since it is not normal memory, like you say, I'm wondering how that is supposed to work.

For now

1. virtio-mem adds all (possible) aligned memory via add_memory() to Linux
2. Requires user space to online the memory blocks / configure a zone.

For 2., only ZONE_NORMAL really works right now and is recommended to use. As you correctly note, that does not give you any guarantees how much memory you can unplug again (e.g, fragmentation with unmovable data), but is good enough for the first version (with focus on memory hotplug, not unplug). ZONE_MOVABLE support is in the works.

However, we cannot blindly expose all memory to ZONE_MOVABLE (zone imbalances leading to rashes), and sometimes also don't want to (e.g., gigantic pages). Without spoilering too much, a mixture would be nice.

Quoted 4 lines

As far as I can tell there would be a lot of inconsistencies in userspace interfaces which provide memory / zone information. Or I'm not getting the point of all of this at all.

All memory/zone stats are properly fixed up (similar to ballooning). The only visible inconsistency that *might* happen when unplugging memory / hotplugging memory in <256MB on s390x, is that the number of memory block devices (/sys/devices/system/memory/...) might indicate more memory than actually available (e.g., via lsmem).

[1]
https://lists.oasis-open.org/archives/virtio-comment/202006/msg00012.html
[2] https://lore.kernel.org/kvm/[email protected]/
[3]
https://events19.linuxfoundation.org/wp-content/uploads/2017/12/virtio-mem-Paravirtualized-Memory-David-Hildenbrand-Red-Hat-1.pdf
[4] https://www.youtube.com/watch?v=H65FDUDPu9s

Signature · 3 lines
Thanks,

David / dhildenb
#29 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 42 lines · David Hildenbrand
Quoted 18 lines · Heiko Carstens
Quoted 17 lines · David Hildenbrand
Quoted 3 lines

So, are you saying that even at IPL time there might already be memory devices attached to the system? And the kernel should _not_ treat them as normal memory?

Sorry if that was unclear. Yes, we can have such devices (including memory areas) on a cold boot/reboot/kexec. In addition, they might pop up at runtime (e.g., hotplugging a virtio-mem device). The device is in charge of exposing that area and deciding what to do with it.

The kernel should never treat them as normal memory (IOW, system RAM). Not during a cold boot, not during a reboot. The device driver is responsible for deciding how to use that memory (e.g., add it as system RAM), and which parts of that memory are actually valid to be used (even if a tprot might succeed it might not be valid to use just yet - I guess somewhat similar to doing a tport on a dcss area - AFAIK, you also don't want to use it like normal memory).

Quoted 7 lines
Quoted 4 lines

E.g., on x86-64, memory exposed via virtio-mem or virtio-pmem is never exposed via the e820 map. The only trace that there might be *something* now/in the future is indicated via ACPI SRAT tables. This takes currently care of indicating the maximum possible PFN.

Ok, but all of this needa to be documented somewhere. This raises a couple of questions to me:

I assume this mostly targets virtio-mem, because the semantics of virtio-mem provided memory are extra-weird (in contrast to rather static virtio-pmem, which is essentially just an emulated NVDIMM - a disk mapped into physical memory).

Regarding documentation (some linked in the cover letter), so far I have (generic/x86-64)

1. https://virtio-mem.gitlab.io/
2. virtio spec proposal [1]
3. QEMU 910b25766b33 ("virtio-mem: Paravirtualized memory hot(un)plug")
4. Linux 5f1f79bbc9 ("virtio-mem: Paravirtualized memory hotplug")
5. Linux cover letter [2]
6. KVM forum talk [3] [4]

Quoted 2 lines

As your questions go quite into technical detail, and I don't feel like rewriting the doc here :) , I suggest looking at [2], 1, and 5.

Sorry, I suggest looking at [3] (not [2]) first. Includes pictures and a comparison to memory ballooning (and DIMM-based memory hotplug).

Quoted 4 lines

[3]
https://events19.linuxfoundation.org/wp-content/uploads/2017/12/virtio-mem-Paravirtualized-Memory-David-Hildenbrand-Red-Hat-1.pdf
[4] https://www.youtube.com/watch?v=H65FDUDPu9s

Signature · 3 lines
Thanks,

David / dhildenb
#30 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 10 lines · David Hildenbrand
Quoted 10 lines

Regarding documentation (some linked in the cover letter), so far I have (generic/x86-64)

1. https://virtio-mem.gitlab.io/
2. virtio spec proposal [1]
3. QEMU 910b25766b33 ("virtio-mem: Paravirtualized memory hot(un)plug")
4. Linux 5f1f79bbc9 ("virtio-mem: Paravirtualized memory hotplug")
5. Linux cover letter [2]
6. KVM forum talk [3] [4]

Quoted 5 lines
Quoted 2 lines

As your questions go quite into technical detail, and I don't feel like rewriting the doc here :) , I suggest looking at [2], 1, and 5.

Sorry, I suggest looking at [3] (not [2]) first. Includes pictures and a comparison to memory ballooning (and DIMM-based memory hotplug).

Ok, thanks for the pointers!

So I would go for what you suggested with option 2: provide a new diagnose which tells the kernel where the memory device area is (probably just start + size?), and leave all other interfaces alone.

This looks to me like the by far "cleanest" solution which does not add semantics to existing interfaces, where it is questionable if this wouldn't cause problems in the future.

#31 lore raw

Re: [PATCH RFC 2/5] s390x: implement diag260

Quoted 11 lines · Heiko Carstens
Quoted 10 lines · David Hildenbrand
Quoted 10 lines

Regarding documentation (some linked in the cover letter), so far I have (generic/x86-64)

1. https://virtio-mem.gitlab.io/
2. virtio spec proposal [1]
3. QEMU 910b25766b33 ("virtio-mem: Paravirtualized memory hot(un)plug")
4. Linux 5f1f79bbc9 ("virtio-mem: Paravirtualized memory hotplug")
5. Linux cover letter [2]
6. KVM forum talk [3] [4]

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

As your questions go quite into technical detail, and I don't feel like rewriting the doc here :) , I suggest looking at [2], 1, and 5.

Sorry, I suggest looking at [3] (not [2]) first. Includes pictures and a comparison to memory ballooning (and DIMM-based memory hotplug).

Ok, thanks for the pointers!

Thanks for having a look. Once the s390x part is in good shape, I'll add proper documentation (+spec updates regarding exact system reset handling on s390x).

Quoted 4 lines

So I would go for what you suggested with option 2: provide a new diagnose which tells the kernel where the memory device area is (probably just start + size?), and leave all other interfaces alone.

Ha, that's precisely what I hacked previously today :) Have a new diag500 ("KVM hypercall") subcode (4) to give start+size of the area reserved for memory devices. Will send a new RFC this week to showcase how it would look like.

Quoted 4 lines

This looks to me like the by far "cleanest" solution which does not add semantics to existing interfaces, where it is questionable if this wouldn't cause problems in the future.

Yes, same thoughts over here!

Signature · 3 lines
Thanks,

David / dhildenb
#32 lore raw

RFC PATCH 3/5 s390x: prepare device memory address space

In reply to David Hildenbrand

Let's allocate the device memory information and setup the device memory address space. Expose the maximum ramsize via SCLP and the actual initial ramsize via diag260.

---
 hw/s390x/s390-virtio-ccw.c         | 43 ++++++++++++++++++++++++++++++
 hw/s390x/sclp.c                    | 12 +++++++--
 include/hw/s390x/s390-virtio-ccw.h |  3 +++
 target/s390x/diag.c                |  4 +--
 4 files changed, 58 insertions(+), 4 deletions(-)
hw/s390x/s390-virtio-ccw.c +43 -0
index 2e6d292c23..577590e623 100644
@@ -160,6 +160,35 @@ static void virtio_ccw_register_hcalls(void)
                                    virtio_ccw_hcall_early_printk);
 }
 
+static void s390_device_memory_init(MachineState *machine)
+{
+    MemoryRegion *sysmem = get_system_memory();
+
+    machine->device_memory = g_malloc0(sizeof(*machine->device_memory));
+
+    /* initialize device memory address space */
+    if (machine->ram_size < machine->maxram_size) {
+        ram_addr_t device_mem_size = machine->maxram_size - machine->ram_size;
+
+        if (QEMU_ALIGN_UP(machine->maxram_size, MiB) != machine->maxram_size) {
+            error_report("maximum memory size must by aligned to 1 MB");
+            exit(EXIT_FAILURE);
+        }
+
+        machine->device_memory->base = machine->ram_size;
+        if (machine->device_memory->base + device_mem_size < device_mem_size) {
+            error_report("unsupported amount of maximum memory: " RAM_ADDR_FMT,
+                         machine->maxram_size);
+            exit(EXIT_FAILURE);
+        }
+
+        memory_region_init(&machine->device_memory->mr, OBJECT(machine),
+                           "device-memory", device_mem_size);
+        memory_region_add_subregion(sysmem, machine->device_memory->base,
+                                    &machine->device_memory->mr);
+    }
+}
+
 static void s390_memory_init(MachineState *machine)
 {
     MemoryRegion *sysmem = get_system_memory();
@@ -194,6 +223,11 @@ static void s390_memory_init(MachineState *machine)
     s390_skeys_init();
     /* Initialize storage attributes device */
     s390_stattrib_init();
+
+    /* Support for memory devices is glued to compat machines. */
+    if (memory_devices_allowed()) {
+        s390_device_memory_init(machine);
+    }
 }
 
 static void s390_init_ipl_dev(const char *kernel_filename,
@@ -617,6 +651,7 @@ static void ccw_machine_class_init(ObjectClass *oc, void *data)
     s390mc->cpu_model_allowed = true;
     s390mc->css_migration_enabled = true;
     s390mc->hpage_1m_allowed = true;
+    s390mc->memory_devices_allowed = true;
     mc->init = ccw_init;
     mc->reset = s390_machine_reset;
     mc->hot_add_cpu = s390_hot_add_cpu;
@@ -713,6 +748,11 @@ bool hpage_1m_allowed(void)
     return get_machine_class()->hpage_1m_allowed;
 }
 
+bool memory_devices_allowed(void)
+{
+    return get_machine_class()->memory_devices_allowed;
+}
+
 static char *machine_get_loadparm(Object *obj, Error **errp)
 {
     S390CcwMachineState *ms = S390_CCW_MACHINE(obj);
@@ -831,8 +871,11 @@ static void ccw_machine_5_0_instance_options(MachineState *machine)
 
 static void ccw_machine_5_0_class_options(MachineClass *mc)
 {
+    S390CcwMachineClass *s390mc = S390_MACHINE_CLASS(mc);
+
     ccw_machine_5_1_class_options(mc);
     compat_props_add(mc->compat_props, hw_compat_5_0, hw_compat_5_0_len);
+    s390mc->memory_devices_allowed = false;
 }
 DEFINE_CCW_MACHINE(5_0, "5.0", false);
 
hw/s390x/sclp.c +10 -2
index f59195e15a..85d3505597 100644
@@ -22,6 +22,7 @@
 #include "hw/s390x/event-facility.h"
 #include "hw/s390x/s390-pci-bus.h"
 #include "hw/s390x/ipl.h"
+#include "hw/s390x/s390-virtio-ccw.h"
 
 static inline SCLPDevice *get_sclp_device(void)
 {
@@ -110,8 +111,15 @@ static void read_SCP_info(SCLPDevice *sclp, SCCB *sccb)
         read_info->rnsize2 = cpu_to_be32(rnsize);
     }
 
-    /* we don't support standby memory, maxram_size is never exposed */
-    rnmax = machine->ram_size >> sclp->increment_size;
+    /*
+     * Support for maxram was added with support for memory devices. The
+     * size of the initial memory is exposed via diag260.
+     */
+    if (memory_devices_allowed()) {
+        rnmax = machine->maxram_size >> sclp->increment_size;
+    } else {
+        rnmax = machine->ram_size >> sclp->increment_size;
+    }
     if (rnmax < 0x10000) {
         read_info->rnmax = cpu_to_be16(rnmax);
     } else {
include/hw/s390x/s390-virtio-ccw.h +3 -0
index cd1dccc6e3..3a1e7e2a6d 100644
@@ -41,6 +41,7 @@ typedef struct S390CcwMachineClass {
     bool cpu_model_allowed;
     bool css_migration_enabled;
     bool hpage_1m_allowed;
+    bool memory_devices_allowed;
 } S390CcwMachineClass;
 
 /* runtime-instrumentation allowed by the machine */
@@ -49,6 +50,8 @@ bool ri_allowed(void);
 bool cpu_model_allowed(void);
 /* 1M huge page mappings allowed by the machine */
 bool hpage_1m_allowed(void);
+/* Allow memory devices and diag260. */
+bool memory_devices_allowed(void);
 
 /**
  * Returns true if (vmstate based) migration of the channel subsystem
target/s390x/diag.c +2 -2
index c3b1e24b2c..6b33eb0efc 100644
@@ -32,8 +32,8 @@ void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
     ram_addr_t addr, length;
     uint64_t tmp;
 
-    /* TODO: Unlock with new QEMU machine. */
-    if (false) {
+    /* Support for diag260 is glued to support for memory devices. */
+    if (!memory_devices_allowed()) {
         s390_program_interrupt(env, PGM_OPERATION, ra);
         return;
     }
Signature · 1 line
2.26.2
#33 lore raw

Re: [PATCH RFC 3/5] s390x: prepare device memory address space

Quoted 11 lines · On Wed, 8 Jul 2020 20:51:33 +0200 David Hildenbrand <[email protected]> wrote:

Let's allocate the device memory information and setup the device memory address space. Expose the maximum ramsize via SCLP and the actual initial ramsize via diag260.

---
 hw/s390x/s390-virtio-ccw.c         | 43 ++++++++++++++++++++++++++++++
 hw/s390x/sclp.c                    | 12 +++++++--
 include/hw/s390x/s390-virtio-ccw.h |  3 +++
 target/s390x/diag.c                |  4 +--
 4 files changed, 58 insertions(+), 4 deletions(-)

(...)

target/s390x/diag.c void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra) · line 35

Quoted 11 lines
target/s390x/diag.c +1 -2
index c3b1e24b2c..6b33eb0efc 100644
@@ -32,8 +32,8 @@ void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
     ram_addr_t addr, length;
     uint64_t tmp;
 
-    /* TODO: Unlock with new QEMU machine. */
-    if (false) {
+    /* Support for diag260 is glued to support for memory devices. */

I'm wondering why you need to do this... sure, the availability of a new diagnose could be perceived as a guest-visible change, but does the information presented change anything? Without memory devices, it will just duplicate the information already reported via SCLP, IIUC?

Quoted 4 lines
target/s390x/diag.c +1 -0
+    if (!memory_devices_allowed()) {
         s390_program_interrupt(env, PGM_OPERATION, ra);
         return;
     }
#34 lore raw

Re: [PATCH RFC 3/5] s390x: prepare device memory address space

Quoted 25 lines · Cornelia Huck
Quoted 11 lines · On Wed, 8 Jul 2020 20:51:33 +0200 David Hildenbrand <[email protected]> wrote:

Let's allocate the device memory information and setup the device memory address space. Expose the maximum ramsize via SCLP and the actual initial ramsize via diag260.

---
 hw/s390x/s390-virtio-ccw.c         | 43 ++++++++++++++++++++++++++++++
 hw/s390x/sclp.c                    | 12 +++++++--
 include/hw/s390x/s390-virtio-ccw.h |  3 +++
 target/s390x/diag.c                |  4 +--
 4 files changed, 58 insertions(+), 4 deletions(-)

(...)

Quoted 8 lines
target/s390x/diag.c +0 -0
index c3b1e24b2c..6b33eb0efc 100644
@@ -32,8 +32,8 @@ void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra)
     ram_addr_t addr, length;
     uint64_t tmp;
 
Quoted 8 lines

target/s390x/diag.c void handle_diag_260(CPUS390XState *env, uint64_t r1, uint64_t r3, uintptr_t ra) · line 35

Quoted 3 lines
target/s390x/diag.c +1 -2
-    /* TODO: Unlock with new QEMU machine. */
-    if (false) {
+    /* Support for diag260 is glued to support for memory devices. */

I'm wondering why you need to do this... sure, the availability of a new diagnose could be perceived as a guest-visible change, but does the information presented change anything? Without memory devices, it will just duplicate the information already reported via SCLP, IIUC?

Yes, it's essentially providing redundant information without memory devices.

One could sense diag260 in the guest and assume it will work on successive invocations. E.g., issue subcode 0xc while checking for exceptions, then issue subcode 0x10 without checking for exceptions. If we migrate in between, we could be in trouble.

Yes, it's somewhat unlikely, I don't have a strong opinion here. Gluing it to some migration-safe mechanism (here, the machine) felt like the right thing to do.

Signature · 3 lines
Thanks,

David / dhildenb
#35 lore raw

RFC PATCH 4/5 s390x: implement virtio-mem-ccw

In reply to David Hildenbrand

Add a proper CCW proxy device, similar to the PCI variant.

---
 hw/s390x/virtio-ccw-mem.c | 165 ++++++++++++++++++++++++++++++++++++++
 hw/s390x/virtio-ccw.h     |  13 +++
 2 files changed, 178 insertions(+)
 create mode 100644 hw/s390x/virtio-ccw-mem.c
hw/s390x/virtio-ccw-mem.c new file +165 -0
new file mode 100644
index 0000000000..ae856d7ad4
@@ -0,0 +1,165 @@
+/*
+ * Virtio MEM CCW device
+ *
+ * Copyright (C) 2020 Red Hat, Inc.
+ *
+ * Authors:
+ *  David Hildenbrand <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include "hw/qdev-properties.h"
+#include "hw/virtio/virtio.h"
+#include "qapi/error.h"
+#include "qemu/module.h"
+#include "virtio-ccw.h"
+#include "hw/mem/memory-device.h"
+#include "qapi/qapi-events-misc.h"
+
+static void virtio_ccw_mem_realize(VirtioCcwDevice *ccw_dev, Error **errp)
+{
+    VirtIOMEMCcw *ccw_mem = VIRTIO_MEM_CCW(ccw_dev);
+    DeviceState *vdev = DEVICE(&ccw_mem->vdev);
+
+    qdev_realize(vdev, BUS(&ccw_dev->bus), errp);
+}
+
+static void virtio_ccw_mem_set_addr(MemoryDeviceState *md, uint64_t addr,
+                                    Error **errp)
+{
+    object_property_set_uint(OBJECT(md), addr, VIRTIO_MEM_ADDR_PROP, errp);
+}
+
+static uint64_t virtio_ccw_mem_get_addr(const MemoryDeviceState *md)
+{
+    return object_property_get_uint(OBJECT(md), VIRTIO_MEM_ADDR_PROP,
+                                    &error_abort);
+}
+
+static MemoryRegion *virtio_ccw_mem_get_memory_region(MemoryDeviceState *md,
+                                                      Error **errp)
+{
+    VirtIOMEMCcw *ccw_mem = VIRTIO_MEM_CCW(md);
+    VirtIOMEM *vmem = VIRTIO_MEM(&ccw_mem->vdev);
+    VirtIOMEMClass *vmc = VIRTIO_MEM_GET_CLASS(vmem);
+
+    return vmc->get_memory_region(vmem, errp);
+}
+
+static uint64_t virtio_ccw_mem_get_plugged_size(const MemoryDeviceState *md,
+                                                Error **errp)
+{
+    return object_property_get_uint(OBJECT(md), VIRTIO_MEM_SIZE_PROP,
+                                    errp);
+}
+
+static void virtio_ccw_mem_fill_device_info(const MemoryDeviceState *md,
+                                            MemoryDeviceInfo *info)
+{
+    VirtioMEMDeviceInfo *vi = g_new0(VirtioMEMDeviceInfo, 1);
+    VirtIOMEMCcw *ccw_mem = VIRTIO_MEM_CCW(md);
+    VirtIOMEM *vmem = VIRTIO_MEM(&ccw_mem->vdev);
+    VirtIOMEMClass *vpc = VIRTIO_MEM_GET_CLASS(vmem);
+    DeviceState *dev = DEVICE(md);
+
+    if (dev->id) {
+        vi->has_id = true;
+        vi->id = g_strdup(dev->id);
+    }
+
+    /* let the real device handle everything else */
+    vpc->fill_device_info(vmem, vi);
+
+    info->u.virtio_mem.data = vi;
+    info->type = MEMORY_DEVICE_INFO_KIND_VIRTIO_MEM;
+}
+
+static void virtio_ccw_mem_size_change_notify(Notifier *notifier, void *data)
+{
+    VirtIOMEMCcw *ccw_mem = container_of(notifier, VirtIOMEMCcw,
+                                         size_change_notifier);
+    DeviceState *dev = DEVICE(ccw_mem);
+    const uint64_t * const size_p = data;
+    const char *id = NULL;
+
+    if (dev->id) {
+        id = g_strdup(dev->id);
+    }
+
+    qapi_event_send_memory_device_size_change(!!id, id, *size_p);
+}
+
+static void virtio_ccw_mem_instance_init(Object *obj)
+{
+    VirtIOMEMCcw *ccw_mem = VIRTIO_MEM_CCW(obj);
+    VirtIOMEMClass *vmc;
+    VirtIOMEM *vmem;
+
+    virtio_instance_init_common(obj, &ccw_mem->vdev, sizeof(ccw_mem->vdev),
+                                TYPE_VIRTIO_MEM);
+
+    ccw_mem->size_change_notifier.notify = virtio_ccw_mem_size_change_notify;
+    vmem = VIRTIO_MEM(&ccw_mem->vdev);
+    vmc = VIRTIO_MEM_GET_CLASS(vmem);
+    /*
+     * We never remove the notifier again, as we expect both devices to
+     * disappear at the same time.
+     */
+    vmc->add_size_change_notifier(vmem, &ccw_mem->size_change_notifier);
+
+    object_property_add_alias(obj, VIRTIO_MEM_BLOCK_SIZE_PROP,
+                              OBJECT(&ccw_mem->vdev),
+                              VIRTIO_MEM_BLOCK_SIZE_PROP);
+    object_property_add_alias(obj, VIRTIO_MEM_SIZE_PROP, OBJECT(&ccw_mem->vdev),
+                              VIRTIO_MEM_SIZE_PROP);
+    object_property_add_alias(obj, VIRTIO_MEM_REQUESTED_SIZE_PROP,
+                              OBJECT(&ccw_mem->vdev),
+                              VIRTIO_MEM_REQUESTED_SIZE_PROP);
+}
+
+static Property virtio_ccw_mem_properties[] = {
+    DEFINE_PROP_BIT("ioeventfd", VirtioCcwDevice, flags,
+                    VIRTIO_CCW_FLAG_USE_IOEVENTFD_BIT, true),
+    DEFINE_PROP_UINT32("max_revision", VirtioCcwDevice, max_rev,
+                       VIRTIO_CCW_MAX_REV),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void virtio_ccw_mem_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtIOCCWDeviceClass *k = VIRTIO_CCW_DEVICE_CLASS(klass);
+    MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(klass);
+
+    k->realize = virtio_ccw_mem_realize;
+    device_class_set_props(dc, virtio_ccw_mem_properties);
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+
+    mdc->get_addr = virtio_ccw_mem_get_addr;
+    mdc->set_addr = virtio_ccw_mem_set_addr;
+    mdc->get_plugged_size = virtio_ccw_mem_get_plugged_size;
+    mdc->get_memory_region = virtio_ccw_mem_get_memory_region;
+    mdc->fill_device_info = virtio_ccw_mem_fill_device_info;
+}
+
+static const TypeInfo virtio_ccw_mem = {
+    .name = TYPE_VIRTIO_MEM_CCW,
+    .parent = TYPE_VIRTIO_CCW_DEVICE,
+    .instance_size = sizeof(VirtIOMEMCcw),
+    .instance_init = virtio_ccw_mem_instance_init,
+    .class_init = virtio_ccw_mem_class_init,
+    .interfaces = (InterfaceInfo[]) {
+        { TYPE_MEMORY_DEVICE },
+        { }
+    },
+};
+
+static void virtio_ccw_mem_register(void)
+{
+    type_register_static(&virtio_ccw_mem);
+}
+
+type_init(virtio_ccw_mem_register)
hw/s390x/virtio-ccw.h +13 -0
index c0e3355248..77aa87c41f 100644
@@ -29,6 +29,7 @@
 #endif /* CONFIG_VHOST_VSOCK */
 #include "hw/virtio/virtio-gpu.h"
 #include "hw/virtio/virtio-input.h"
+#include "hw/virtio/virtio-mem.h"
 
 #include "hw/s390x/s390_flic.h"
 #include "hw/s390x/css.h"
@@ -256,4 +257,16 @@ typedef struct VirtIOInputHIDCcw {
     VirtIOInputHID vdev;
 } VirtIOInputHIDCcw;
 
+/* virtio-mem-ccw */
+
+#define TYPE_VIRTIO_MEM_CCW "virtio-mem-ccw"
+#define VIRTIO_MEM_CCW(obj) \
+        OBJECT_CHECK(VirtIOMEMCcw, (obj), TYPE_VIRTIO_MEM_CCW)
+
+typedef struct VirtIOMEMCcw {
+    VirtioCcwDevice parent_obj;
+    VirtIOMEM vdev;
+    Notifier size_change_notifier;
+} VirtIOMEMCcw;
+
 #endif
Signature · 1 line
2.26.2
#36 lore raw

Re: [PATCH RFC 4/5] s390x: implement virtio-mem-ccw

Quoted 8 lines · On Wed, 8 Jul 2020 20:51:34 +0200 David Hildenbrand <[email protected]> wrote:

Add a proper CCW proxy device, similar to the PCI variant.

---
 hw/s390x/virtio-ccw-mem.c | 165 ++++++++++++++++++++++++++++++++++++++
 hw/s390x/virtio-ccw.h     |  13 +++
 2 files changed, 178 insertions(+)
 create mode 100644 hw/s390x/virtio-ccw-mem.c

(...)

hw/s390x/virtio-ccw-mem.c line 100

Quoted 6 lines
hw/s390x/virtio-ccw-mem.c new file +6 -0
+static void virtio_ccw_mem_instance_init(Object *obj)
+{
+    VirtIOMEMCcw *ccw_mem = VIRTIO_MEM_CCW(obj);
+    VirtIOMEMClass *vmc;
+    VirtIOMEM *vmem;
+

I think you want

    ccw_dev->force_revision_1 = true;

here (similar to forcing virtio-pci to modern-only.)

Quoted 11 lines
hw/s390x/virtio-ccw-mem.c new file +11 -0
+    virtio_instance_init_common(obj, &ccw_mem->vdev, sizeof(ccw_mem->vdev),
+                                TYPE_VIRTIO_MEM);
+
+    ccw_mem->size_change_notifier.notify = virtio_ccw_mem_size_change_notify;
+    vmem = VIRTIO_MEM(&ccw_mem->vdev);
+    vmc = VIRTIO_MEM_GET_CLASS(vmem);
+    /*
+     * We never remove the notifier again, as we expect both devices to
+     * disappear at the same time.
+     */
+    vmc->add_size_change_notifier(vmem, &ccw_mem->size_change_notifier);

hw/s390x/virtio-ccw-mem.c line 121

Quoted 10 lines
hw/s390x/virtio-ccw-mem.c new file +10 -0
+
+    object_property_add_alias(obj, VIRTIO_MEM_BLOCK_SIZE_PROP,
+                              OBJECT(&ccw_mem->vdev),
+                              VIRTIO_MEM_BLOCK_SIZE_PROP);
+    object_property_add_alias(obj, VIRTIO_MEM_SIZE_PROP, OBJECT(&ccw_mem->vdev),
+                              VIRTIO_MEM_SIZE_PROP);
+    object_property_add_alias(obj, VIRTIO_MEM_REQUESTED_SIZE_PROP,
+                              OBJECT(&ccw_mem->vdev),
+                              VIRTIO_MEM_REQUESTED_SIZE_PROP);
+}

(...)

(have not looked at the rest yet)

#37 lore raw

Re: [PATCH RFC 4/5] s390x: implement virtio-mem-ccw

Quoted 23 lines · Cornelia Huck
Quoted 8 lines · On Wed, 8 Jul 2020 20:51:34 +0200 David Hildenbrand <[email protected]> wrote:

Add a proper CCW proxy device, similar to the PCI variant.

---
 hw/s390x/virtio-ccw-mem.c | 165 ++++++++++++++++++++++++++++++++++++++
 hw/s390x/virtio-ccw.h     |  13 +++
 2 files changed, 178 insertions(+)
 create mode 100644 hw/s390x/virtio-ccw-mem.c

(...)

hw/s390x/virtio-ccw-mem.c line 100

Quoted 6 lines
hw/s390x/virtio-ccw-mem.c new file +6 -0
+static void virtio_ccw_mem_instance_init(Object *obj)
+{
+    VirtIOMEMCcw *ccw_mem = VIRTIO_MEM_CCW(obj);
+    VirtIOMEMClass *vmc;
+    VirtIOMEM *vmem;
+

I think you want

Quoted 3 lines
    ccw_dev->force_revision_1 = true;

here (similar to forcing virtio-pci to modern-only.)

Ah, that's the magic bit, was looking for that. Thanks!

Signature · 3 lines
Thanks,

David / dhildenb
#38 lore raw

RFC PATCH 5/5 s390x: initial support for virtio-mem

In reply to David Hildenbrand

Let's wire up the initial, basic virtio-mem implementation in QEMU. It will have to see some important extensions (esp., resizeable allocations) before it can be considered production ready. Also, the focus on the Linux driver side is on memory hotplug, there are a lot of things optimize in the future to improve memory unplug capabilities. However, the basics are in place.

Block migration for now, as we'll have to take proper care of storage keys and storage attributes. Also, make sure to not hotplug huge pages to a setup without huge pages.

With a Linux guest that supports virtio-mem (and has CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE set for now), a basic example.

1. Start a VM with 2G initial memory and a virtio-mem device with a maximum capacity of 18GB (and an initial size of 300M):

    sudo qemu-system-s390x \
        --enable-kvm \
        -m 2G,maxmem=20G \
        -smp 4 \
        -nographic \
        -chardev socket,id=monitor,path=/var/tmp/monitor,server,nowait \
        -mon chardev=monitor,mode=readline \
        -net nic -net user \
        -hda s390x.cow2 \
        -object memory-backend-ram,id=mem0,size=18G \
        -device virtio-mem-ccw,id=vm0,memdev=mem0,requested-size=300M

2. Query the current size of virtio-mem device:

    (qemu) info memory-devices
    Memory device [virtio-mem]: "vm0"
      memaddr: 0x80000000
      node: 0
      requested-size: 314572800
      size: 314572800
      max-size: 19327352832
      block-size: 1048576
      memdev: /objects/mem0

3. Request to grow it to 8GB:

    (qemu) qom-set vm0 requested-size 8G
    (qemu) info memory-devices
    Memory device [virtio-mem]: "vm0"
      memaddr: 0x80000000
      node: 0
      requested-size: 8589934592
      size: 8589934592
      max-size: 19327352832
      block-size: 1048576
      memdev: /objects/mem0

4. Request to shrink it to 800M (might take a while, might not fully succeed, and might not be able to remove memory blocks in Linux):
(qemu) qom-set vm0 requested-size 800M
(qemu) info memory-devices
Memory device [virtio-mem]: "vm0"

    memaddr: 0x80000000
    node: 0
    requested-size: 838860800
    size: 838860800
    max-size: 19327352832
    block-size: 1048576
    memdev: /objects/mem0

Note: Due to lack of resizeable allocations, we will go ahead and reserve a 18GB vmalloc area + size the QEMU RAM slot + KVM mamory slot 18GB. echo 1 > /proc/sys/vm/overcommit_memory might be required for now. In the future, this area will instead grow on actual demand and shrink when possible.

---
 hw/s390x/Kconfig           |   1 +
 hw/s390x/Makefile.objs     |   1 +
 hw/s390x/s390-virtio-ccw.c | 116 ++++++++++++++++++++++++++++++++++++-
 hw/virtio/virtio-mem.c     |   2 +
 4 files changed, 118 insertions(+), 2 deletions(-)
hw/s390x/Kconfig +1 -0
index 5e7d8a2bae..b8619c1adc 100644
@@ -10,3 +10,4 @@ config S390_CCW_VIRTIO
     select SCLPCONSOLE
     select VIRTIO_CCW
     select MSI_NONBROKEN
+    select VIRTIO_MEM_SUPPORTED
hw/s390x/Makefile.objs +1 -0
index a46a1c7894..924775d6f0 100644
@@ -20,6 +20,7 @@ obj-$(CONFIG_VIRTIO_NET) += virtio-ccw-net.o
 obj-$(CONFIG_VIRTIO_BLK) += virtio-ccw-blk.o
 obj-$(call land,$(CONFIG_VIRTIO_9P),$(CONFIG_VIRTFS)) += virtio-ccw-9p.o
 obj-$(CONFIG_VHOST_VSOCK) += vhost-vsock-ccw.o
+obj-$(CONFIG_VIRTIO_MEM) += virtio-ccw-mem.o
 endif
 obj-y += css-bridge.o
 obj-y += ccw-device.o
hw/s390x/s390-virtio-ccw.c +114 -2

151 lines not on this page · raw message

hw/virtio/virtio-mem.c +2 -0
index 65850530e7..e1b3275089 100644
@@ -53,6 +53,8 @@
  */
 #if defined(TARGET_X86_64) || defined(TARGET_I386)
 #define VIRTIO_MEM_USABLE_EXTENT (2 * (128 * MiB))
+#elif defined(TARGET_S390X)
+#define VIRTIO_MEM_USABLE_EXTENT (2 * (256 * MiB))
 #else
 #error VIRTIO_MEM_USABLE_EXTENT not defined
 #endif
Signature · 1 line
2.26.2