The memory region reference is increased when insert a range
into flatview range array, then decreased by destroy flatview.
If some flat range merged by flatview_simplify, the memory region
reference can not be decreased by destroy flatview any more.
In this case, start virtual machine by the command line:
qemu-system-x86_64
-name guest=ubuntu,debug-threads=on
-machine pc,accel=kvm,usb=off,dump-guest-core=off
-cpu host
-m 16384
-realtime mlock=off
-smp 8,sockets=2,cores=4,threads=1
-object memory-backend-file,id=ram-node0,prealloc=yes,mem-path=/dev/hugepages,share=yes,size=8589934592
-numa node,nodeid=0,cpus=0-3,memdev=ram-node0
-object memory-backend-file,id=ram-node1,prealloc=yes,mem-path=/dev/hugepages,share=yes,size=8589934592
-numa node,nodeid=1,cpus=4-7,memdev=ram-node1
-no-user-config
-nodefaults
-rtc base=utc
-no-shutdown
-boot strict=on
-device piix3-usb-uhci,id=usb,bus=pci.0,addr=0x1.0x2
-device virtio-scsi-pci,id=scsi0,bus=pci.0,addr=0x2
-device virtio-serial-pci,id=virtio-serial0,bus=pci.0,addr=0x3
-drive file=ubuntu.qcow2,format=qcow2,if=none,id=drive-virtio-disk0,cache=none,aio=native
-device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x4,drive=drive-virtio-disk0,id=virtio-disk0,bootindex=1
-chardev pty,id=charserial0
-device isa-serial,chardev=charserial0,id=serial0
-device usb-tablet,id=input0,bus=usb.0,port=1
-vnc 0.0.0.0:0
-device VGA,id=video0,vgamem_mb=16,bus=pci.0,addr=0x5
-device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x6
-msg timestamp=on
And run the script in guest OS:
while true
do
setpci -s 00:06.0 04.b=03
setpci -s 00:06.0 04.b=07
done
I found the reference of node0 HostMemoryBackendFile is a big one.
(gdb) p numa_info[0]->node_memdev->parent.ref
$6 = 1636278
(gdb)
Backports commit 838ec1177c45812ff177a35ff93990e9eb7f70c3 from qemu
In the M-profile architecture, when we do a vector table fetch and it
fails, we need to report a HardFault. Whether this is a Secure HF or
a NonSecure HF depends on several things. If AIRCR.BFHFNMINS is 0
then HF is always Secure, because there is no NonSecure HardFault.
Otherwise, the answer depends on whether the 'underlying exception'
(MemManage, BusFault, SecureFault) targets Secure or NonSecure. (In
the pseudocode, this is handled in the Vector() function: the final
exc.isSecure is calculated by looking at the exc.isSecure from the
exception returned from the memory access, not the isSecure input
argument.)
We weren't doing this correctly, because we were looking at
the target security domain of the exception we were trying to
load the vector table entry for. This produces errors of two kinds:
* a load from the NS vector table which hits the "NS access
to S memory" SecureFault should end up as a Secure HardFault,
but we were raising an NS HardFault
* a load from the S vector table which causes a BusFault
should raise an NS HardFault if BFHFNMINS == 1 (because
in that case all BusFaults are NonSecure), but we were raising
a Secure HardFault
Correct the logic.
We also fix a comment error where we claimed that we might
be escalating MemManage to HardFault, and forgot about SecureFault.
(Vector loads can never hit MPU access faults, because they're
always aligned and always use the default address map.)
Backports commit 51c9122e92b776a3f16af0b9282f1dc5012e2a19 from qemu
The ARMv5 architecture didn't specify detailed per-feature ID
registers. Now that we're using the MVFR0 register fields to
gate the existence of VFP instructions, we need to set up
the correct values in the cpu->isar structure so that we still
provide an FPU to the guest.
This fixes a regression in the arm926 and arm1026 CPUs, which
are the only ones that both have VFP and are ARMv5 or earlier.
This regression was introduced by the VFP refactoring, and more
specifically by commits 1120827fa182f0e76 and 266bd25c485597c,
which accidentally disabled VFP short-vector support and
double-precision support on these CPUs.
Backports commit cb7cef8b32033f6284a47d797edd5c19c5491698 from qemu
When we converted to using feature bits in 602f6e42cfbf we missed out
the fact (dp && arm_dc_feature(s, ARM_FEATURE_V8)) was supported for
-cpu max configurations. This caused a regression in the GCC test
suite. Fix this by setting the appropriate bits in mvfr1.FPHP to
report ARMv8-A with FP support (but not ARMv8.2-FP16).
Fixes: https://bugs.launchpad.net/qemu/+bug/1836078
Backports commit 45b1a243b81a7c9ae56235937280711dd9914ca7 from qemu
We have some potential race conditions vs our user-exec signal
handler that will be solved with this barrier.
Backports commit 359896dfa4e9707e1acea99129d324250fccab04 from qemu
This patch fixes two problems:
(1) The inputs to the EXTR insn were reversed,
(2) The input constraints use rZ, which means that we need to use
the REG0 macro in order to supply XZR for a constant 0 input.
Fixes: 464c2969d5d
Backports commit 1789d4274b851fb8fdf4a947ce5474c63e813d0d from qemu
On a 64-bit host, discard any replications of the 32-bit
sign bit when performing the shift and merge.
Backports commit 80f4d7c3ae216c191fb403e149bcba88d6aa40bb from qemu
This operation can always be emitted, even if we need to
fall back to xor. Adjust the assertions to match.
Backports commit 11978f6f58f1d3d66429f7ff897524f693d823ce from qemu
In commit e9d652824b0 we extracted the vfp_set_fpscr_to_host()
function but failed at calling it in the correct place, we call
it after xregs[ARM_VFP_FPSCR] is modified.
Fix by calling this function before we update FPSCR.
Backports commit 85795187f416326f87177cabc39fae1911f04c50 from qemu
Off by one error in the EL2 and EL3 tests. Remove the test
against EL3 entirely, since it must always be true.
Backports commit 6a02a73211c5bc634fccd652777230954b83ccba from qemu
Add new version of Cascadelake-Server CPU model, setting
stepping=5 and enabling the IA32_ARCH_CAPABILITIES MSR
with some flags.
The new feature will introduce a new host software requirement,
breaking our CPU model runnability promises. This means we can't
enable the new CPU model version by default in QEMU 4.1, because
management software isn't ready yet to resolve CPU model aliases.
This is why "pc-*-4.1" will keep returning Cascadelake-Server-v1
if "-cpu Cascadelake-Server" is specified.
Includes a test case to ensure the right combinations of
machine-type + CPU model + command-line feature flags will work
as expected.
Backports commit fd63c6d1a5f77d689ee06f6561677c012a988223 from qemu
The old CPU models will be just aliases for specific versions of
the original CPU models.
Backports commit 53db89d93bebe70a3e7f4c45933deffcf3e7cb62 from qemu
Add versions of CPU models that are equivalent to their -IBRS,
-noTSX and -IBRS variants.
The separate variants will eventually be removed and become
aliases for these CPU versions.
Backports commit d86a708815c3bec0b934760e6bdab7eb647087b8 from qemu
Add support for registration of multiple versions of CPU models.
The existing CPU models will be registered with a "-v1" suffix.
The -noTSX, -IBRS, and -IBPB CPU model variants will become
versions of the original models in a separate patch, so
make sure we register no versions for them.
Backports commit dcafd1ef0af227ef87f7a6dec8fc66d7d2e2442d from qemu
The CPUID.1F as Intel V2 Extended Topology Enumeration Leaf would be
exposed if guests want to emulate multiple software-visible die within
each package. Per Intel's SDM, the 0x1f is a superset of 0xb, thus they
can be generated by almost same code as 0xb except die_offset setting.
If the number of dies per package is greater than 1, the cpuid_min_level
would be adjusted to 0x1f regardless of whether the host supports CPUID.1F.
Likewise, the CPUID.1F wouldn't be exposed if env->nr_dies < 2.
Backports commit a94e1428991f741e2c6636e7c8df7f8d1905d983 from qemu
In new sockets/dies/cores/threads model, the apicid of logical cpu could
imply die level info of guest cpu topology thus x86_apicid_from_cpu_idx()
need to be refactored with #dies value, so does apicid_*_offset().
To keep semantic compatibility, the legacy pkg_offset which helps to
generate CPUIDs such as 0x3 for L3 cache should be mapping to die_offset.
Backports commit d65af288a84d8bf8c27e55d45545f52f016c08a7 from qemu
The field die_id (default as 0) and has_die_id are introduced to X86CPU.
Following the legacy smp check rules, the die_id validity is added to
the same contexts as leagcy smp variables such as hmp_hotpluggable_cpus(),
machine_set_cpu_numa_node(), cpu_slot_to_string() and pc_cpu_pre_plug().
Backports relevant bits from 176d2cda0dee9f4f78f604ad72d6a111e8e38f3b
from qemu
The die-level as the first PC-specific cpu topology is added to the leagcy
cpu topology model, which has one die per package implicitly and only the
numbers of sockets/cores/threads are configurable.
In the new model with die-level support, the total number of logical
processors (including offline) on board will be calculated as:
\#cpus = #sockets * #dies * #cores * #threads
and considering compatibility, the default value for #dies would be
initialized to one in x86_cpu_initfn() and pc_machine_initfn().
Backports commit c26ae610811e8d52f4fc73e3ae0a8bc4a24d6763 from qemu
Coverity points out (CID 1402195) that the loop in trans_VMOV_imm_dp()
that iterates over the destination registers in a short-vector VMOV
accidentally throws away the returned updated register number
from vfp_advance_dreg(). Add the missing assignment. (We got this
correct in trans_VMOV_imm_sp().)
Backports commit 89a11ff756410aecb87d2c774df6e45dbf4105c1 from qemu
Thumb instructions in an IT block are set up to be conditionally
executed depending on a set of condition bits encoded into the IT
bits of the CPSR/XPSR. The architecture specifies that if the
condition bits are 0b1111 this means "always execute" (like 0b1110),
not "never execute"; we were treating it as "never execute". (See
the ConditionHolds() pseudocode in both the A-profile and M-profile
Arm ARM.)
This is a bit of an obscure corner case, because the only legal
way to get to an 0b1111 set of condbits is to do an exception
return which sets the XPSR/CPSR up that way. An IT instruction
which encodes a condition sequence that would include an 0b1111 is
UNPREDICTABLE, and for v8A the CONSTRAINED UNPREDICTABLE choices
for such an IT insn are to NOP, UNDEF, or treat 0b1111 like 0b1110.
Add a comment noting that we take the latter option.
Backports commit 5529de1e5512c05276825fa8b922147663fd6eac from qemu
In the various helper functions for v7M/v8M instructions, use
the _ra versions of cpu_stl_data() and friends. Otherwise we
may get wrong behaviour or an assert() due to not being able
to locate the TB if there is an exception on the memory access
or if it performs an IO operation when in icount mode
Backports commit 2884fbb60412049ec92389039ae716b32057382e from qemu
In preparation for supporting TCG disablement on ARM, we move most
of TCG related v7m/v8m helpers and APIs into their own file.
Note: It is easier to review this commit using the 'histogram'
diff algorithm:
$ git diff --diff-algorithm=histogram ...
or
$ git diff --histogram ...
Backports commit 7aab5a8c8bb525ea390b4ebc17ab82c0835cfdb6 from qemu
Semihosting hooks either SVC or HLT instructions, and inside KVM
both of those go to EL1, ie to the guest, and can't be trapped to
KVM.
Let check_for_semihosting() return False when not running on TCG.
backports commit 91f78c58da9ba78c8ed00f5d822b701765be8499 from qemu
We make a few sub-directories recursively, in particular
$(TARGET_DIRS).
For goal "all", we do it the nice way: "all" has a prerequisite
subdir-T for each T in $(TARGET_DIRS), and T's recipe runs make
recursively. Behaves nicely with -j and -k.
For other goals such as "clean" and "install", the recipe runs make
recursively in a for loop. Ignores -j and -k.
The next commit will fix that for "clean" and "install". This commit
prepares the ground by renaming the targets we use for "all" to
include the goal for the sub-make. This will permit reusing them for
goals other than "all".
Targets subdir-T for T in $(TARGET_DIRS) run "make all" in T. Rename
to T/all, and declare phony.
Targets romsubdir-R for R in $(ROMS) run "make" in pc-bios/R. Default
goal is "all" for all R. Rename to pc-bios/R/all, and declare phony.
The remainder are renamed just for consistency.
Target subdir-dtc runs "make libbft/libfdt.a" in dtc. Rename to
dtc/all, and declare phony.
Target subdir-capstone runs make $(BUILD_DIR)/capstone/$(LIBCAPSTONE)
in $(SRC_PATH)/capstone. Rename to capstone/all, and declare phony.
Target subdir-slirp runs "make" in $(SRC_PATH)/slirp. Default goal is
all, which builds $(BUILD_DIR)/libslirp.a. Rename to slirp/all, and
declare phony.
Backports commit 3b8593eeaa5778ae118f0cc2837e615acd13baeb from qemu
When commit bdf523e6923 made configure generate config.status, it
added a fallback to Makefile to smooth the transition, with a TODO
"code can be removed after QEMU 1.7." It's been more than five years.
Remove it.
Backports commit cdb69b08f984b7cca2acb1989c8eae79390a23f3 from qemu
This prevents a load reservation from being placed in one context/process,
then being used in another, resulting in an SC succeeding incorrectly and
breaking atomics.
Backports commit c13b169f1a3dd158d6c75727cdc388f95988db39 from qemu
The various CSR instructions have been split out of the base ISA as part
of the ratification process. This patch adds a Zicsr argument, which
disables all the CSR instructions.
Backports commit 591bddea8d874e1500921de0353818e5586618f5 from qemu
fence.i has been split out of the base ISA as part of the ratification
process. This patch adds a Zifencei argument, which disables the
fence.i instruction.
Backports commit 50fba816cd226001bec3e495c39879deb2fa5432 from qemu
Set the priv spec version 1.11.0 as the default and allow selecting it
via the command line.
Backports commit e3147506b02edcdd7c14ebb41a10fcc3027dcc5c from qemu
1.11 defines mcountinhibit, which has the same numeric CSR value as
mucounteren from 1.09.1 but has different semantics. This patch enables
the CSR for 1.11-based targets, which is trivial to implement because
the counters in QEMU never tick (legal according to the spec).
Backports commit 747a43e818dc36bd50ef98c2b11a7c31ceb810fa from qemu
Restructure the deprecated CPUs to make it clear in the code that these
are depreated. They are already marked as deprecated in
qemu-deprecated.texi. There are no functional changes.
Backports commit c1fb65e63cfca4506a14b084afd0eca2dc464fe8 from qemu
The current implementation returns 1 (PMP check success) if the address is in
range even if the PMP entry is off. This is a bug.
For example, if there is a PMP check in S-Mode which is in range, but its PMP
entry is off, this will succeed, which it should not.
The patch fixes this bug by only checking the PMP permissions if the address is
in range and its corresponding PMP entry it not off. Otherwise, it will keep
the ret = -1 which will be checked and handled correctly at the end of the
function.
Backports commit f8162068f18f2f264a0355938784f54089234211 from qemu
The PMP should be checked when doing a page table walk, and report access
fault exception if the to-be-read PTE failed the PMP check.
Backports commit 1f447aec787bfbbd078afccae44fc4c92acb4fed from qemu
The current PMP check function checks for env->priv which is not the effective
memory privilege mode.
For example, mstatus.MPRV could be set while executing in M-Mode, and in that
case the privilege mode for the PMP check should be S-Mode rather than M-Mode
(in env->priv) if mstatus.MPP == PRV_S.
This patch passes the effective memory privilege mode to the PMP check.
Functions that call the PMP check should pass the correct memory privilege mode
after reading mstatus' MPRV/MPP or hstatus.SPRV (if Hypervisor mode exists).
Backports commit cc0fdb298517ce56c770803447f8b02a90271152 from qemu
Section 3.6 in RISC-V v1.10 privilege specification states that PMP violations
report "access exceptions." The current PMP implementation has
a bug which wrongly reports "page exceptions" on PMP violations.
This patch fixes this bug by reporting the correct PMP access exceptions
trap values.
Backports commit 635b0b0ea39a13d1a3df932452e5728aebbb3f6e from qemu
The current implementation unnecessarily checks for PMP even if MMU translation
failed. This may trigger a wrong PMP access exception instead of
a page exception.
For example, the very first instruction fetched after the first satp write in
S-Mode will trigger a PMP access fault instead of an instruction fetch page
fault.
This patch prioritises MMU exceptions over PMP exceptions and only checks for
PMP if MMU translation succeeds. This patch is required for future commits
that properly report PMP exception violations if PTW succeeds.
Backports commit e0f8fa72deba7ac7a7ae06ba25e6498aaad93ace from qemu
This patch adds support for the riscv_cpu_unassigned_access call
and will raise a load or store access fault.
Backports commit cbf5827693addaff4e4d2102afedbf078a204eb2 from qemu
A wrong address is passed to `pmp_is_in_range` while checking if a
memory access is within a PMP range.
Since the ending address of the pmp range (i.e., pmp_state.addr[i].ea)
is set to the last address in the range (i.e., pmp base + pmp size - 1),
memory accesses containg the last address in the range will always fail.
For example, assume that a PMP range is 4KB from 0x87654000 such that
the last address within the range is 0x87654fff.
1-byte access to 0x87654fff should be considered to be fully inside the
PMP range.
However the access now fails and complains partial inclusion because
pmp_is_in_range(env, i, addr + size) returns 0 whereas
pmp_is_in_range(env, i, addr) returns 1.
Backports commit 49db9fa1fd7c252596b53cf80876e06f407d09ed from qemu
