Rather than setting cpu->as unconditionally in cpu_exec_init
(and then having target-i386 override this later), don't set
it until the first call to cpu_address_space_init.
This requires us to initialise the address space for
both TCG and KVM (KVM doesn't need the AS listener but
it does require cpu->as to be set).
For target CPUs which don't set up any address spaces (currently
everything except i386), add the default address_space_memory
in qemu_init_vcpu().
Backports commit 56943e8cc14b7eeeab67d1942fa5d8bcafe3e53f from qemu
For inbound migration we really want to be able to set the PSR without
having any side effects, but cpu_put_psr() calls cpu_check_irqs() which
might try to deliver CPU interrupts. Split cpu_put_psr() into the
no-side-effect and side-effect parts.
This includes reordering the cpu_check_irqs() to the end of cpu_put_psr(),
because that function may actually end up calling cpu_interrupt(), which
does not seem like a good thing to happen in the middle of updating the PSR.
Backports commit 4552a09dd4055c806b7df8c595dc0fb8951834be from qemu
arm_regime_using_lpae_format checks whether the LPAE extension is used
for stage 1 translation regimes. MMU indexes not exclusively of a stage 1
regime won't work with this method.
In case of ARMMMUIdx_S12NSE0 or ARMMMUIdx_S12NSE1, offset these values
by ARMMMUIdx_S1NSE0 to get the right index indicating a stage 1
translation regime.
Rename also the function to arm_s1_regime_using_lpae_format and update
the comments to reflect the change.
Backports commit deb2db996cbb9470b39ae1e383791ef34c4eb3c2 from qemu
memcpy can take a large amount of time for small reads and writes.
Handle the common case of reading s/g descriptors from memory (there
is no corresponding "write" case that is as common, because writes
often use address_space_st* functions) by inlining the relevant
parts of address_space_read into the caller.
Backports commit 3cc8f884996584630734a90c9b3c535af81e3c92 from qemu
We want to inline the case where there is only one iteration, because
then the compiler can also inline the memcpy. As a start, extract
everything after the first address_space_translate call.
Backports commit a203ac702e0720135fac8b1f2061d119814c1798 from qemu
Replace qemu_ram_free_from_ptr() with qemu_ram_free().
The only difference between qemu_ram_free_from_ptr() and
qemu_ram_free() is that g_free_rcu() is used instead of
call_rcu(reclaim_ramblock). We can safely replace it because:
* RAM blocks allocated by qemu_ram_alloc_from_ptr() always have
RAM_PREALLOC set;
* reclaim_ramblock(block) will do nothing except g_free(block)
if RAM_PREALLOC is set at block->flags.
Backports commit a29ac16632aec6065c72985b9f7eeb1ca6fbef4a from qemu
Qemu does not generally perform alignment checks. However, the ARM ARM
requires implementation of alignment exceptions for a number of cases
including LDREX, and Windows-on-ARM relies on this.
This change adds plumbing to enable alignment checks on loads using
MO_ALIGN, a do_unaligned_access hook to raise the exception (data
abort), and uses the new aligned loads in LDREX (for all but
single-byte loads).
Backports commit 30901475b91ef1f46304404ab4bfe89097f61b96 from qemu
Add a function to find a RAMBlock by name; use it in two
of the places that already open code that loop; we've
got another use later in postcopy.
Backports commit e3dd74934f2d2c8c67083995928ff68e8c1d0030 from qemu
Postcopy sends RAMBlock names and offsets over the wire (since it can't
rely on the order of ramaddr being the same), and it starts out with
HVA fault addresses from the kernel.
qemu_ram_block_from_host translates a HVA into a RAMBlock, an offset
in the RAMBlock and the global ram_addr_t value.
Rewrite qemu_ram_addr_from_host to use qemu_ram_block_from_host.
Provide qemu_ram_get_idstr since its the actual name text sent on the
wire.
Backports commit 422148d3e56c3c9a07c0cf36c1e0a0b76f09c357 from qemu
This makes ROM blocks resizeable. This infrastructure is required for other
functionality we have queued.
Backports commit aaf03019175949eda5087329448b8a0033b89479 from qemu
Add Performance Counter (4) and XNP (5) register numbers to RDHWR.
Add check_hwrena() to simplify access control checkings.
Add RDHWR support to microMIPS R6.
Backports commit b00c72180c36510bf9b124e190bd520e3b7e1358 from qemu
A QEMU breakpoint match is not definitely an architectural breakpoint
match. If an exception is generated unconditionally during translation,
it is hardly possible to ignore it in the debug exception handler.
Generate a call to a helper to check CPU breakpoints and raise an
exception only if any breakpoint matches architecturally.
Backports commit 5d98bf8f38c17a348ab6e8af196088cd4953acd0 from qemu
It is no longer used, so tidy up everything reached by it.
This includes the gen_opc_* arrays, the search_pc parameter
and the inline gen_intermediate_code_internal functions.
Backports commit 4e5e1215156662b2b153255c49d4640d82c5568b from qemu
This patch improves exception handling in MIPS.
Instructions generate several types of exceptions.
When exception is generated, it breaks the execution of the current
translation block. Implementation of the exceptions handling does not
correctly restore icount for the instruction which caused the exception.
In most cases icount will be decreased by the value equal to the size of
TB. This patch passes pointer to the translation block internals to the
exception handler. It allows correct restoring of the icount value.
Backports commit 9c708c7f9fbb813a3fac02f2728e51e62f2f5ffc from qemu
This patch introduces loop exit function, which also
restores guest CPU state according to the value of host
program counter.
Backports commit 1c3c8af1fb40a481c07749e0448644d9b7700415 from qemu
This is particularly useful when we abort in error_propagate(),
because there the stack backtrace doesn't lead to where the error was
created. Looks like this:
Unexpected error in parse_block_error_action() at .../qemu/blockdev.c:322:
qemu-system-x86_64: -drive if=none,werror=foo: 'foo' invalid write error action
Aborted (core dumped)
Note: to get this example output, I monkey-patched drive_new() to pass
&error_abort to blockdev_init().
To keep the error handling boiler plate from growing even more, all
error_setFOO() become macros expanding into error_setFOO_internal()
with additional __FILE__, __LINE__, __func__ arguments. Not exactly
pretty, but it works.
The macro trickery breaks down when you take the address of an
error_setFOO(). Fortunately, we do that in just one place: qemu-ga's
Windows VSS provider and requester DLL wants to call
error_setg_win32() through a function pointer "to avoid linking glib
to the DLL". Use error_setg_win32_internal() there. The use of the
function pointer is already wrapped in a macro, so the churn isn't
bad.
Code size increases by some 35KiB for me (0.7%). Tolerable. Could be
less if we passed relative rather than absolute source file names to
the compiler, or forwent reporting __func__.
Backports commit 1e9b65bb1bad51735cab6c861c29b592dccabf0e from qemu
For the A64 instruction set, the semihosting call instruction
is 'HLT 0xf000'. Wire this up to call do_arm_semihosting()
if semihosting is enabled.
Backports commit 8012c84ff92a36d05dfe61af9b24dd01a7ea25e4 from qemu
Guest CPU TLB maintenance operations may be sufficiently
specialized to only need to flush TLB entries corresponding
to a particular MMU index. Implement cputlb functions for
this, to avoid the inefficiency of flushing TLB entries
which we don't need to.
Backports commit d7a74a9d4a68e27b3a8ceda17bb95cb0a23d8e4d from qemu
On CPUs with EL3, there are two physical timers, one for Secure and one
for Non-secure. Implement this extra timer and the AArch64 registers
which access it.
Backports commit b4d3978c2fdf944e428a46d2850dbd950b6fbe78 from qemu
Introduce a new crypto/ directory that will (eventually) contain
all the cryptographic related code. This initially defines a
wrapper for initializing gnutls and for computing hashes with
gnutls. The former ensures that gnutls is guaranteed to be
initialized exactly once in QEMU regardless of CLI args. The
block quorum code currently fails to initialize gnutls so it
only works by luck, if VNC server TLS is not requested. The
hash APIs avoids the need to litter the rest of the code with
preprocessor checks and simplifies callers by allocating the
correct amount of memory for the requested hash.
Backports commit ddbb0d09661f5fce21b335ba9aea8202d189b98e from qemu
This introduces the memory region property "global_locking". It is true
by default. By setting it to false, a device model can request BQL-free
dispatching of region accesses to its r/w handlers. The actual BQL
break-up will be provided in a separate patch.
Backports commit 196ea13104f802c508e57180b2a0d2b3418989a3 from qemu
Currently we use DISAS_WFE for both WFE and YIELD instructions.
This is functionally correct because at the moment both of them
are implemented as "yield this CPU back to the top level loop so
another CPU has a chance to run". However it's rather confusing
that YIELD ends up calling HELPER(wfe), and if we ever want to
implement real behaviour for WFE and SEV it's likely to trip us up.
Split out the yield codepath to use DISAS_YIELD and a new
HELPER(yield) function, and have HELPER(wfe) call HELPER(yield).
Backports commit 049e24a191c212d9468db84169197887f2c91586 from qemu
ERETNC is identical to ERET except that an ERETNC will not clear the LLbit
that is set by execution of an LL instruction, and thus when placed between
an LL and SC sequence, will never cause the SC to fail.
Presence of ERETNC is denoted by the Config5.LLB.
Backports commit ce9782f40ac16660ea9437bfaa2c9c34d5ed8110 from qemu
MIPS SIMD Architecture vector loads and stores require misalignment support.
MSA Memory access should work as an atomic operation. Therefore, it has to
check validity of all addresses for a vector store access if it is spanning
into two pages.
Separating helper functions for each data format as format is known in
translation.
To use mmu_idx from cpu_mmu_index() instead of calculating it from hflag.
Removing save_cpu_state() call in translation because it is able to use
cpu_restore_state() on fault as GETRA() is passed.
Backports commit adc370a48fd26b92188fa4848dfb088578b1936c from qemu
These include page table walks, SVM accesses and SMM state save accesses.
The bulk of the patch is obtained with
sed -i 's/\(\<[a-z_]*_phys\(_notdirty\)\?\>(cs\)->as,/x86_\1,/'
Backports commit b216aa6c0fcbaa8ff4128969c14594896a5485a4 from qemu
The cpu_physical_memory_reset_dirty() function is sometimes used
together with cpu_physical_memory_get_dirty(). This is not atomic since
two separate accesses to the dirty memory bitmap are made.
Turn cpu_physical_memory_reset_dirty() and
cpu_physical_memory_clear_dirty_range_type() into the atomic
cpu_physical_memory_test_and_clear_dirty().
Backports commit 03eebc9e3246b9b3f5925aa41f7dfd7c1e467875 from qemu
While it is obvious that cpu_physical_memory_get_dirty returns true even if
a single page is dirty, the same is not true for cpu_physical_memory_get_clean;
one would expect that it returns true only if all the pages are clean, but
it actually looks for even one clean page. (By contrast, the caller of that
function, cpu_physical_memory_range_includes_clean, has a good name).
To clarify, rename the function to cpu_physical_memory_all_dirty and return
true if _all_ the pages are dirty. This is the opposite of the previous
meaning, because "all are 1" is the same as "not (any is 0)", so we have to
modify cpu_physical_memory_range_includes_clean as well
Backports commit 72b47e79cef36ed6ffc718f10e21001d7ec2a66f from qemu
These days modification of the TLB is done in notdirty_mem_write,
so the virtual address and env pointer as unnecessary.
The new name of the function, tlb_unprotect_code, is consistent with
tlb_protect_code.
Backports commit 9564f52da7eb061326956ed9a468935e3352512d from qemu
For now memory regions only track DIRTY_MEMORY_VGA individually, but
this will change soon. To support this, split memory_region_is_logging
in two functions: one that returns a given bit from dirty_log_mask,
and one that returns the entire mask. memory_region_is_logging gets an
extra parameter so that the compiler flags misuse.
While VGA-specific users (including the Xen listener!) will want to keep
checking that bit, KVM and vhost check for "any bit except migration"
(because migration is handled via the global start/stop listener
callbacks).
Backports commit 2d1a35bef0ed96b3f23535e459c552414ccdbafd from qemu
Updated the interrupt handling to utilize and report through the target EL
exception field. This includes consolidating and cleaning up code where
needed. Target EL is now calculated once in arm_cpu_exec_interrupt() and
do_interrupt was updated to use the target_el exception field. The
necessary code from arm_excp_target_el() was merged in where needed and the
function removed.
Backports commit 012a906b19e99b126403ff4a257617dab9b34163 from qemu
Move the code which sets exception information out of
arm_cpu_handle_mmu_fault and into tlb_fill. tlb_fill
is the only caller which wants to raise_exception()
so it makes more sense for it to handle the whole of
the exception setup.
As part of this cleanup, move the user-mode-only
implementation function for the handle_mmu_fault CPU
method into cpu.c so we don't need to make it globally
visible, and rename the softmmu-only utility function
arm_cpu_handle_mmu_fault to arm_tlb_fill so it's clear
that it's not the same thing.
Backports commit 8c6084bf10fe721929ca94cf16acd6687e61d3ec from qemu
Add new address_space_ld*/st* functions which allow transaction
attributes and error reporting for basic load and stores. These
are named to be in line with the address_space_read/write/rw
buffer operations.
The existing ld/st*_phys functions are now wrappers around
the new functions.
Backports commit 500131154d677930fce35ec3a6f0b5a26bcd2973 from qemu
Add a MemTxAttrs field to the IOTLB, and allow target-specific
code to set it via a new tlb_set_page_with_attrs() function;
pass the attributes through to the device when making IO accesses.
Backports commit fadc1cbe85c6b032d5842ec0d19d209f50fcb375 from qemu
Rather than retaining io_mem_read/write as simple wrappers around
the memory_region_dispatch_read/write functions, make the latter
public and change all the callers to use them, since we need to
touch all the callsites anyway to add MemTxAttrs and MemTxResult
support. Delete io_mem_read and io_mem_write entirely.
(All the callers currently pass MEMTXATTRS_UNSPECIFIED
and convert the return value back to bool or ignore it.)
Backports commit 3b6434953934e6d4a776ed426d8c6d6badee176f from qemu
This for now is a simple TLB flush. This can change later for two
reasons:
1) an AddressSpaceDispatch will be cached in the CPUState object
2) it will not be possible to do tlb_flush once the TCG-generated code
runs outside the BQL.
Backports commit 76e5c76f2e2e0d20bab2cd5c7a87452f711654fb from qemu
Add AArch32 to AArch64 register sychronization functions.
Replace manual register synchronization with new functions in
aarch64_cpu_do_interrupt() and HELPER(exception_return)().
Backports commit ce02049dbf1828b4bc77d921b108a9d84246e5aa from qemu
When EL3 is running in AArch32 (or ARMv7 with Security Extensions)
PAR has a secure and a non-secure instance.
Backports commit 01c097f7960b330c4bf038d34bae17ad6c1ba499 from qemu
Adds a dedicated function and a lookup table for determining the target
exception level of IRQ and FIQ exceptions. The lookup table is taken from the
ARMv7 and ARMv8 specification exception routing tables.
Backports commit 0eeb17d618361a0f4faddc160e33598b23da6dd5 from qemu
Fix CP0.Config3.ISAOnExc write accesses on microMIPS processors. This
bit is mandatory for any processor that implements the microMIPS
instruction set. This bit is r/w for processors that implement both the
standard MIPS and the microMIPS instruction set. This bit is r/o and
hardwired to 1 if only the microMIPS instruction set is implemented.
There is no other bit ever writable in CP0.Config3 so defining a
corresponding `CP0_Config3_rw_bitmask' member in `CPUMIPSState' is I
think an overkill. Therefore make the ability to write the bit rely on
the presence of ASE_MICROMIPS set in the instruction flags.
The read-only case of the microMIPS instruction set being implemented
only can be added when we add support for such a configuration. We do
not currently have such support, we have no instruction flag that would
control the presence of the standard MIPS instruction set nor any
associated code in instruction decoding.
This change is needed to boot a microMIPS Linux kernel successfully,
otherwise it hangs early on as interrupts are enabled and then the
exception handler invoked loops as its first instruction is interpreted
in the wrong execution mode and triggers another exception right away.
And then over and over again.
We already check the current setting of the CP0.Config3.ISAOnExc in
`set_hflags_for_handler' to set the ISA bit correctly on the exception
handler entry so it is the ability to set it that is missing only.
Backports commit 90f12d735d66ac1196d9a2bced039a432eefc03d from qemu
Split arm_gen_test_cc into 3 functions, so that it can be reused
for non-branch TCG comparisons.
Backports commit 6c2c63d3a02c79e9035ca0370cc549d0f938a4dd from qemu
Rather than allow arbitrary shift+trunc, only concern ourselves
with low and high parts. This is all that was being used anyway.
Backports commit 609ad70562793937257c89d07bf7c1370b9fc9aa from qemu
Rather reserving space in the op stream for optimization,
let the optimizer add ops as necessary.
Backports commit a4ce099a7a4b4734c372f6bf28f3362e370f23c1 from qemu
With the linked list scheme we need not leave nops in the stream
that we need to process later.
Backports commit 0c627cdca20155753a536c51385abb73941a59a0 from qemu
Some of these functions are really quite large. We have a number of
things that ought to be circularly dependent, but we duplicated code
to break that chain for the inlines.
This saved 25% of the code size of one of the translators I examined.
This commit fixes the following issues:
- Any unmapped/free'd memory regions (MemoryRegion instances) are not
removed from the object property linked list of its owner (which is
always qdev_get_machine(uc)). This issue makes adding new memory
mapping by calling mem_map() or mem_map_ptr() slower as more and more
memory pages are mapped and unmapped - yes, even if those memory pages
are unmapped, they still impact the speed of future memory page
mappings due to this issue.
- FlatView is not reconstructed after a memory region is freed during
unmapping, which leads to a use-after-free the next time a new memory
region is mapped in address_space_update_topology().
ARM and probably the rest of the arches have significant memory leaks as
they have no release interface.
Additionally, DrMemory does not have 64-bit support and thus I can't
test the 64-bit version under Windows. Under Linux valgrind supports
both 32-bit and 64-bit but there are different macros and code for Linux
and Windows.
