[Qemu-devel] [PULL 0/8] ppc-for-2.12 queue 20180319

[Qemu-devel] [PULL 0/8] ppc-for-2.12 queue 20180319

[David Gibson]

The following changes since commit e1e44a9916b4318e943aecd669e096222cb3eaeb:

  Merge remote-tracking branch 'remotes/xtensa/tags/20180316-xtensa' into staging (2018-03-17 14:15:03 +0000)

are available in the Git repository at:

  git://github.com/dgibson/qemu.git tags/ppc-for-2.12-20180319

for you to fetch changes up to 91c60f124e682a78c9a2ef951e8e58ebab6441d0:

  target/ppc: fix tlbsync to check privilege level depending on GTSE (2018-03-18 21:03:20 +1100)

----------------------------------------------------------------
ppc patch queue for 2018-03-19

This pull request supersedes the one for 2018-03-15.  The only
difference is one patch is removed, since it exposed some code which
triggers ubsan warnings.

Here's the set of accumulated patches now that we're into soft freeze.
I've split new functionality into a ppc-for-2.13 branch, so this only
has bugfixes.  Well.. and a couple of simple cleanups to make bugfixes
easier, some test improvements and a trivial change to make command
line options more obvious.  I think those are all acceptable for soft
freeze.

----------------------------------------------------------------
BALATON Zoltan (2):
      sii3112: Remove unneeded exit function
      ppc440_pcix: Change some error_report to qemu_log_mask(LOG_UNIMP, ...)

Cédric Le Goater (1):
      target/ppc: fix tlbsync to check privilege level depending on GTSE

David Engraf (1):
      PPC e500: Fix gap between u-boot and kernel

Thomas Huth (4):
      tests/boot-serial: Check the 40p machine, too
      hw/ppc/prep: Fix implicit creation of "-drive if=scsi" devices
      hw/misc/macio: Mark the macio devices with user_creatable = false
      hw/ppc/spapr: Allow "spapr-vlan" as NIC model name beside "ibmveth"

 hw/ide/sii3112.c         | 12 ------------
 hw/misc/macio/macio.c    |  2 ++
 hw/ppc/e500.c            |  9 ++++-----
 hw/ppc/ppc440_pcix.c     | 11 +++++++----
 hw/ppc/prep.c            |  2 +-
 hw/ppc/spapr.c           |  5 +++--
 hw/scsi/lsi53c895a.c     |  7 +++++++
 include/hw/pci/pci.h     |  1 +
 target/ppc/translate.c   |  9 +++++++--
 tests/boot-serial-test.c |  6 ++++--
 10 files changed, 36 insertions(+), 28 deletions(-)

[Qemu-devel] [PULL 1/8] sii3112: Remove unneeded exit function

[David Gibson]

From: BALATON Zoltan <balaton@eik.bme.hu>

An exit function was mistakenly left here but it's not needed because
the PCI bars are organised differently in this device. Calling this
exit function during device_del was causing an abort with
memory_region_del_subregion: `Assertion subregion->container == mr' failed.

Reported-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 hw/ide/sii3112.c | 12 ------------
 1 file changed, 12 deletions(-)

diff --git a/hw/ide/sii3112.c b/hw/ide/sii3112.c
index e3896c65b4..743a50ed51 100644
--- a/hw/ide/sii3112.c
+++ b/hw/ide/sii3112.c
@@ -327,17 +327,6 @@ static void sii3112_pci_realize(PCIDevice *dev, Error **errp)
     qemu_register_reset(sii3112_reset, s);
 }
 
-static void sii3112_pci_exitfn(PCIDevice *dev)
-{
-    PCIIDEState *d = PCI_IDE(dev);
-    int i;
-
-    for (i = 0; i < 2; ++i) {
-        memory_region_del_subregion(&d->bmdma_bar, &d->bmdma[i].extra_io);
-        memory_region_del_subregion(&d->bmdma_bar, &d->bmdma[i].addr_ioport);
-    }
-}
-
 static void sii3112_pci_class_init(ObjectClass *klass, void *data)
 {
     DeviceClass *dc = DEVICE_CLASS(klass);
@@ -348,7 +337,6 @@ static void sii3112_pci_class_init(ObjectClass *klass, void *data)
     pd->class_id = PCI_CLASS_STORAGE_RAID;
     pd->revision = 1;
     pd->realize = sii3112_pci_realize;
-    pd->exit = sii3112_pci_exitfn;
     dc->desc = "SiI3112A SATA controller";
     set_bit(DEVICE_CATEGORY_STORAGE, dc->categories);
 }
-- 
2.14.3

[Qemu-devel] [PULL 2/8] tests/boot-serial: Check the 40p machine, too

[David Gibson]

From: Thomas Huth <thuth@redhat.com>

The "40p" machine is using the Open Hack'Ware BIOS, just like the "prep"
machine, so we can test it accordingly with the boot-serial tester, too.
While we're at it, also change the strings that we are using for the
"prep" machine, so that this test now also checks some CLI parameters.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Hervé Poussineau <hpoussin@reactos.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 tests/boot-serial-test.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/tests/boot-serial-test.c b/tests/boot-serial-test.c
index ece25c694f..5b24cd26c1 100644
--- a/tests/boot-serial-test.c
+++ b/tests/boot-serial-test.c
@@ -75,11 +75,13 @@ typedef struct testdef {
 static testdef_t tests[] = {
     { "alpha", "clipper", "", "PCI:" },
     { "ppc", "ppce500", "", "U-Boot" },
-    { "ppc", "prep", "", "Open Hack'Ware BIOS" },
+    { "ppc", "prep", "-m 96", "Memory size: 96 MB" },
+    { "ppc", "40p", "-boot d", "Booting from device d" },
     { "ppc", "g3beige", "", "PowerPC,750" },
     { "ppc", "mac99", "", "PowerPC,G4" },
     { "ppc64", "ppce500", "", "U-Boot" },
-    { "ppc64", "prep", "", "Open Hack'Ware BIOS" },
+    { "ppc64", "prep", "-boot e", "Booting from device e" },
+    { "ppc64", "40p", "-m 192", "Memory size: 192 MB" },
     { "ppc64", "mac99", "", "PowerPC,970FX" },
     { "ppc64", "pseries", "", "Open Firmware" },
     { "ppc64", "powernv", "-cpu POWER8", "OPAL" },
-- 
2.14.3

[Qemu-devel] [PULL 3/8] hw/ppc/prep: Fix implicit creation of "-drive if=scsi" devices

[David Gibson]

From: Thomas Huth <thuth@redhat.com>

The global hack for creating SCSI devices has recently been removed,
but this apparently broke SCSI devices on some boards that were not
ready for this change yet. For the 40p machine you now get:

$ ppc64-softmmu/qemu-system-ppc64 -M 40p -cdrom x.iso
qemu-system-ppc64: -cdrom x.iso: machine type does not support if=scsi,bus=0,unit=2

Fix it by providing a lsi53c810_create() function that takes care
of calling scsi_bus_legacy_handle_cmdline() after creating the
corresponding SCSI controller.

Fixes: 1454509726719e0933c800fad00d6999752688ea
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 hw/ppc/prep.c        | 2 +-
 hw/scsi/lsi53c895a.c | 7 +++++++
 include/hw/pci/pci.h | 1 +
 3 files changed, 9 insertions(+), 1 deletion(-)

diff --git a/hw/ppc/prep.c b/hw/ppc/prep.c
index 5c78503069..a1e7219db6 100644
--- a/hw/ppc/prep.c
+++ b/hw/ppc/prep.c
@@ -787,7 +787,7 @@ static void ibm_40p_init(MachineState *machine)
         qdev_prop_set_uint32(dev, "equipment", 0xc0);
         qdev_init_nofail(dev);
 
-        pci_create_simple(pci_bus, PCI_DEVFN(1, 0), "lsi53c810");
+        lsi53c810_create(pci_bus, PCI_DEVFN(1, 0));
 
         /* XXX: s3-trio at PCI_DEVFN(2, 0) */
         pci_vga_init(pci_bus);
diff --git a/hw/scsi/lsi53c895a.c b/hw/scsi/lsi53c895a.c
index f3d4c4d230..160657f4b9 100644
--- a/hw/scsi/lsi53c895a.c
+++ b/hw/scsi/lsi53c895a.c
@@ -2279,3 +2279,10 @@ void lsi53c895a_create(PCIBus *bus)
 
     scsi_bus_legacy_handle_cmdline(&s->bus);
 }
+
+void lsi53c810_create(PCIBus *bus, int devfn)
+{
+    LSIState *s = LSI53C895A(pci_create_simple(bus, devfn, "lsi53c810"));
+
+    scsi_bus_legacy_handle_cmdline(&s->bus);
+}
diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
index d8c18c7fa4..e255941b5a 100644
--- a/include/hw/pci/pci.h
+++ b/include/hw/pci/pci.h
@@ -708,6 +708,7 @@ PCIDevice *pci_create(PCIBus *bus, int devfn, const char *name);
 PCIDevice *pci_create_simple(PCIBus *bus, int devfn, const char *name);
 
 void lsi53c895a_create(PCIBus *bus);
+void lsi53c810_create(PCIBus *bus, int devfn);
 
 qemu_irq pci_allocate_irq(PCIDevice *pci_dev);
 void pci_set_irq(PCIDevice *pci_dev, int level);
-- 
2.14.3

[Qemu-devel] [PULL 4/8] hw/misc/macio: Mark the macio devices with user_creatable = false

[David Gibson]

From: Thomas Huth <thuth@redhat.com>

The macio devices currently cause a crash when the user tries to
instantiate them on a different machine:

$ ppc64-softmmu/qemu-system-ppc64 -device macio-newworld
Unexpected error in qemu_chr_fe_init() at chardev/char-fe.c:222:
qemu-system-ppc64: -device macio-newworld: Device 'serial0' is in use
Aborted (core dumped)

These devices are clearly not intended to be creatable by the user
since they are using serial_hds[] directly in their instance_init
function. So let's mark them with user_creatable = false.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 hw/misc/macio/macio.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/hw/misc/macio/macio.c b/hw/misc/macio/macio.c
index af1bd46b4b..454244f59e 100644
--- a/hw/misc/macio/macio.c
+++ b/hw/misc/macio/macio.c
@@ -406,6 +406,8 @@ static void macio_class_init(ObjectClass *klass, void *data)
     k->class_id = PCI_CLASS_OTHERS << 8;
     dc->props = macio_properties;
     set_bit(DEVICE_CATEGORY_BRIDGE, dc->categories);
+    /* Reason: Uses serial_hds in macio_instance_init */
+    dc->user_creatable = false;
 }
 
 static const TypeInfo macio_oldworld_type_info = {
-- 
2.14.3

[Qemu-devel] [PULL 5/8] PPC e500: Fix gap between u-boot and kernel

[David Gibson]

From: David Engraf <david.engraf@sysgo.com>

This patch moves the gap between u-boot and kernel at the correct location.

Signed-off-by: David Engraf <david.engraf@sysgo.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 hw/ppc/e500.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/hw/ppc/e500.c b/hw/ppc/e500.c
index 2238f963c4..9a85a41362 100644
--- a/hw/ppc/e500.c
+++ b/hw/ppc/e500.c
@@ -1009,6 +1009,10 @@ void ppce500_init(MachineState *machine, PPCE500Params *params)
     }
 
     cur_base = loadaddr + payload_size;
+    if (cur_base < (32 * 1024 * 1024)) {
+        /* u-boot occupies memory up to 32MB, so load blobs above */
+        cur_base = (32 * 1024 * 1024);
+    }
 
     /* Load bare kernel only if no bios/u-boot has been provided */
     if (machine->kernel_filename && !kernel_as_payload) {
@@ -1025,11 +1029,6 @@ void ppce500_init(MachineState *machine, PPCE500Params *params)
         cur_base += kernel_size;
     }
 
-    if (cur_base < (32 * 1024 * 1024)) {
-        /* u-boot occupies memory up to 32MB, so load blobs above */
-        cur_base = (32 * 1024 * 1024);
-    }
-
     /* Load initrd. */
     if (machine->initrd_filename) {
         initrd_base = (cur_base + INITRD_LOAD_PAD) & ~INITRD_PAD_MASK;
-- 
2.14.3

[Qemu-devel] [PULL 6/8] hw/ppc/spapr: Allow "spapr-vlan" as NIC model name beside "ibmveth"

[David Gibson]

From: Thomas Huth <thuth@redhat.com>

With the new "--nic" command line parameter option, the "old" way of
specifying a NIC model via the nd_table[] is becoming more prominent
again. But for the pseries "spapr-vlan" device, there is a confusing
discrepancy between the model name that is used for "--device" (i.e.
"spapr-vlan") and the model name that has to be used for "--net nic"
or the new "--nic" parameter (i.e. "ibmveth"). Since "spapr-vlan" is
the "real" name of the device, let's allow "spapr-vlan" to be used
as model name for the nd_table[] entries, too.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 hw/ppc/spapr.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index 032d03423f..fba76abee2 100644
--- a/hw/ppc/spapr.c
+++ b/hw/ppc/spapr.c
@@ -2607,10 +2607,11 @@ static void spapr_machine_init(MachineState *machine)
         NICInfo *nd = &nd_table[i];
 
         if (!nd->model) {
-            nd->model = g_strdup("ibmveth");
+            nd->model = g_strdup("spapr-vlan");
         }
 
-        if (strcmp(nd->model, "ibmveth") == 0) {
+        if (g_str_equal(nd->model, "spapr-vlan") ||
+            g_str_equal(nd->model, "ibmveth")) {
             spapr_vlan_create(spapr->vio_bus, nd);
         } else {
             pci_nic_init_nofail(&nd_table[i], phb->bus, nd->model, NULL);
-- 
2.14.3

[Qemu-devel] [PULL 7/8] ppc440_pcix: Change some error_report to qemu_log_mask(LOG_UNIMP, ...)

[David Gibson]

From: BALATON Zoltan <balaton@eik.bme.hu>

Using log unimp is more appropriate for these messages and this also
silences them by default so they won't clobber make check output when
tests are added for this board.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 hw/ppc/ppc440_pcix.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/hw/ppc/ppc440_pcix.c b/hw/ppc/ppc440_pcix.c
index ab2626a9de..b1307e6477 100644
--- a/hw/ppc/ppc440_pcix.c
+++ b/hw/ppc/ppc440_pcix.c
@@ -21,6 +21,7 @@
 
 #include "qemu/osdep.h"
 #include "qemu/error-report.h"
+#include "qemu/log.h"
 #include "hw/hw.h"
 #include "hw/ppc/ppc.h"
 #include "hw/ppc/ppc4xx.h"
@@ -286,8 +287,9 @@ static void ppc440_pcix_reg_write4(void *opaque, hwaddr addr,
         break;
 
     default:
-        error_report("%s: unhandled PCI internal register 0x%lx", __func__,
-                     (unsigned long)addr);
+        qemu_log_mask(LOG_UNIMP,
+                      "%s: unhandled PCI internal register 0x%"HWADDR_PRIx"\n",
+                      __func__, addr);
         break;
     }
 }
@@ -377,8 +379,9 @@ static uint64_t ppc440_pcix_reg_read4(void *opaque, hwaddr addr,
         break;
 
     default:
-        error_report("%s: invalid PCI internal register 0x%lx", __func__,
-                     (unsigned long)addr);
+        qemu_log_mask(LOG_UNIMP,
+                      "%s: invalid PCI internal register 0x%" HWADDR_PRIx "\n",
+                      __func__, addr);
         val = 0;
     }
 
-- 
2.14.3

[Qemu-devel] [PULL 8/8] target/ppc: fix tlbsync to check privilege level depending on GTSE

[David Gibson]

From: Cédric Le Goater <clg@kaod.org>

tlbsync also needs to check the Guest Translation Shootdown Enable
(GTSE) bit in the Logical Partition Control Register (LPCR) to
determine at which privilege level it is running.

See commit c6fd28fd573d ("target/ppc: Update tlbie to check privilege
level based on GTSE")

Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Suraj Jitindar Singh <sjitindarsingh@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 target/ppc/translate.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/target/ppc/translate.c b/target/ppc/translate.c
index 0a0c090c99..218665b408 100644
--- a/target/ppc/translate.c
+++ b/target/ppc/translate.c
@@ -4526,7 +4526,7 @@ static void gen_tlbie(DisasContext *ctx)
     TCGv_i32 t1;
 
     if (ctx->gtse) {
-        CHK_SV; /* If gtse is set then tblie is supervisor privileged */
+        CHK_SV; /* If gtse is set then tlbie is supervisor privileged */
     } else {
         CHK_HV; /* Else hypervisor privileged */
     }
@@ -4553,7 +4553,12 @@ static void gen_tlbsync(DisasContext *ctx)
 #if defined(CONFIG_USER_ONLY)
     GEN_PRIV;
 #else
-    CHK_HV;
+
+    if (ctx->gtse) {
+        CHK_SV; /* If gtse is set then tlbsync is supervisor privileged */
+    } else {
+        CHK_HV; /* Else hypervisor privileged */
+    }
 
     /* BookS does both ptesync and tlbsync make tlbsync a nop for server */
     if (ctx->insns_flags & PPC_BOOKE) {
-- 
2.14.3

Re: [Qemu-devel] [PULL 0/8] ppc-for-2.12 queue 20180319

[Peter Maydell]

On 19 March 2018 at 00:35, David Gibson <david@gibson.dropbear.id.au> wrote:
> The following changes since commit e1e44a9916b4318e943aecd669e096222cb3eaeb:
>
>   Merge remote-tracking branch 'remotes/xtensa/tags/20180316-xtensa' into staging (2018-03-17 14:15:03 +0000)
>
> are available in the Git repository at:
>
>   git://github.com/dgibson/qemu.git tags/ppc-for-2.12-20180319
>
> for you to fetch changes up to 91c60f124e682a78c9a2ef951e8e58ebab6441d0:
>
>   target/ppc: fix tlbsync to check privilege level depending on GTSE (2018-03-18 21:03:20 +1100)
>
> ----------------------------------------------------------------
> ppc patch queue for 2018-03-19
>
> This pull request supersedes the one for 2018-03-15.  The only
> difference is one patch is removed, since it exposed some code which
> triggers ubsan warnings.
>
> Here's the set of accumulated patches now that we're into soft freeze.
> I've split new functionality into a ppc-for-2.13 branch, so this only
> has bugfixes.  Well.. and a couple of simple cleanups to make bugfixes
> easier, some test improvements and a trivial change to make command
> line options more obvious.  I think those are all acceptable for soft
> freeze.
>

Applied, thanks.

-- PMM