riscv-isa-sim.git
9 years agoRefactor memory access code; add MMIO support
Andrew Waterman [Fri, 25 Sep 2015 01:09:26 +0000 (18:09 -0700)]
Refactor memory access code; add MMIO support

Of course, it doesn't do anything yet.

9 years agoUse enum instead of two bools to denote memory access type
Andrew Waterman [Thu, 24 Sep 2015 21:40:40 +0000 (14:40 -0700)]
Use enum instead of two bools to denote memory access type

9 years agocommit log now correctly prints privilege
Scott Beamer [Wed, 16 Sep 2015 00:43:37 +0000 (17:43 -0700)]
commit log now correctly prints privilege

adopting convention of privilege at time of commit, not after commit

9 years agoZero-extend flw, fmv_s_x instructions
Christopher Celio [Tue, 15 Sep 2015 22:05:11 +0000 (15:05 -0700)]
Zero-extend flw, fmv_s_x instructions

   - This makes Spike more consistent with its zero-extending behavior
     regarding other SP operations when placed into the 64b f-registers.

9 years agoprint out current privilege level (if commit log enabled)
Scott Beamer [Sat, 12 Sep 2015 01:04:37 +0000 (18:04 -0700)]
print out current privilege level (if commit log enabled)

9 years agoprint out commit log (if enabled) for all privilege levels
Scott Beamer [Sat, 12 Sep 2015 01:01:11 +0000 (18:01 -0700)]
print out commit log (if enabled) for all privilege levels

9 years agoSimplify register_base_instructions
Andrew Waterman [Fri, 11 Sep 2015 22:13:52 +0000 (15:13 -0700)]
Simplify register_base_instructions

9 years agoInitialize mstatus.prv1/prv2 to U, not S
Andrew Waterman [Fri, 11 Sep 2015 22:13:26 +0000 (15:13 -0700)]
Initialize mstatus.prv1/prv2 to U, not S

This doesn't really matter, but it matches Rocket.

9 years agoSupport 'G' in ISA strings
Andrew Waterman [Fri, 11 Sep 2015 22:13:08 +0000 (15:13 -0700)]
Support 'G' in ISA strings

It's the same as IMAFD.

9 years agoMerge pull request #32 from riscv/insn-list
Andrew Waterman [Fri, 11 Sep 2015 22:05:32 +0000 (15:05 -0700)]
Merge pull request #32 from riscv/insn-list

Eliminate non-portable sed usage

9 years agoFix non-portable sed commands generating insn_list.h
Albert Ou [Fri, 11 Sep 2015 00:01:59 +0000 (17:01 -0700)]
Fix non-portable sed commands generating insn_list.h

9 years agoImprove instruction fetch
Andrew Waterman [Tue, 8 Sep 2015 22:09:23 +0000 (15:09 -0700)]
Improve instruction fetch

- Performance for variable-length instructions is much better
- Refill is simpler and faster
- Support for instructions with overlapping opcodes (e.g. C.ADD + C.JALR)

9 years agoAdd facility to instrument specific opcodes
Andrew Waterman [Tue, 8 Sep 2015 22:05:31 +0000 (15:05 -0700)]
Add facility to instrument specific opcodes

It's not ideal, because it requires modifying tracer.h.  A more general
facility would allow overriding the instruction execution function for
a given opcode dynamically.

9 years agoRefer to LICENSE in some newer source files
Andrew Waterman [Tue, 8 Sep 2015 22:03:03 +0000 (15:03 -0700)]
Refer to LICENSE in some newer source files

9 years agoMove towards RVC v1.8
Andrew Waterman [Fri, 4 Sep 2015 20:39:42 +0000 (13:39 -0700)]
Move towards RVC v1.8

9 years agoDon't automatically run autoconf
Andrew Waterman [Wed, 2 Sep 2015 20:36:10 +0000 (13:36 -0700)]
Don't automatically run autoconf

Run it yourself if you modify configure.ac, and commit the configure script.

9 years agoMerge pull request #29 from pmundkur/devel
Andrew Waterman [Thu, 6 Aug 2015 22:59:08 +0000 (15:59 -0700)]
Merge pull request #29 from pmundkur/devel

Enable execution log generation in non-interactive mode.

9 years agoAdd an option (-l) to display a log of execution in non-interactive mode.
Prashanth Mundkur [Thu, 6 Aug 2015 21:57:07 +0000 (14:57 -0700)]
Add an option (-l) to display a log of execution in non-interactive mode.

Interactive (-d) mode overrides this option when both are specified.

9 years agoAdded error message when trying to use histogram
Christopher Celio [Thu, 30 Jul 2015 20:08:46 +0000 (13:08 -0700)]
Added error message when trying to use histogram

9 years agoUpdate README.md for freg info
Christopher Celio [Tue, 21 Jul 2015 19:45:17 +0000 (12:45 -0700)]
Update README.md for freg info

9 years agoMerge pull request #28 from sbeamer/master
Andrew Waterman [Wed, 15 Jul 2015 17:57:11 +0000 (12:57 -0500)]
Merge pull request #28 from sbeamer/master

fixes clang warning (#26)

9 years agosimplified default case and added comments
Scott Beamer [Mon, 13 Jul 2015 23:17:13 +0000 (16:17 -0700)]
simplified default case and added comments

9 years agoMerge pull request #27 from sbeamer/master
Andrew Waterman [Sat, 11 Jul 2015 03:16:34 +0000 (22:16 -0500)]
Merge pull request #27 from sbeamer/master

fix clang compile error

9 years agofix clang compile error
Scott Beamer [Sat, 11 Jul 2015 02:38:22 +0000 (19:38 -0700)]
fix clang compile error

9 years agoNew machine-mode timer facility
Andrew Waterman [Sun, 5 Jul 2015 23:47:20 +0000 (16:47 -0700)]
New machine-mode timer facility

9 years agoMerge pull request #25 from vapier/master
Andrew Waterman [Sat, 6 Jun 2015 02:00:16 +0000 (19:00 -0700)]
Merge pull request #25 from vapier/master

few more interactive cleanups and more functionality

9 years agoallow interactive "reg" command to dump all registers
Mike Frysinger [Fri, 5 Jun 2015 13:17:19 +0000 (21:17 +0800)]
allow interactive "reg" command to dump all registers

It you want to scan all the registers at once (or at least a few), having
to dump them one by one is kind of a pain.  Change the behavior so that if
the register number is omitted, it'll dump all of them.

9 years agoadd an interactive "pc" command
Mike Frysinger [Fri, 5 Jun 2015 13:13:04 +0000 (21:13 +0800)]
add an interactive "pc" command

This lets you show the current pc quickly.  Sometimes when displaying
different state you can lose track of what the pc was.  Add a simple
command that plumbs in the existing functions to the user interface.

9 years agounify interactive core processing
Mike Frysinger [Fri, 5 Jun 2015 13:06:52 +0000 (21:06 +0800)]
unify interactive core processing

Different functions in here process the core argument in different ways.
Unify all of them with a utility function.

9 years agoMerge pull request #24 from vapier/master
Andrew Waterman [Thu, 4 Jun 2015 23:06:47 +0000 (16:06 -0700)]
Merge pull request #24 from vapier/master

improve interactive behavior a bit more

9 years agoadd aliases for common interactive functions
Mike Frysinger [Thu, 4 Jun 2015 16:04:52 +0000 (00:04 +0800)]
add aliases for common interactive functions

Rather than using one letter commands everywhere, add support for "run"
and "quit" and "help" as aliases to "r" and "q" and "h" respectively.
This makes it a little inconsistent when some commands are one letters
but others are full words.

9 years agoadd a help screen to interactive mode
Mike Frysinger [Thu, 4 Jun 2015 15:59:00 +0000 (23:59 +0800)]
add a help screen to interactive mode

There's no documentation that covers all the available functionality
(in source or the accompanied docs).  Start a help command so that the
info is always live and up-to-date for people.

9 years agoMerge pull request #23 from vapier/master
Andrew Waterman [Wed, 3 Jun 2015 21:00:07 +0000 (14:00 -0700)]
Merge pull request #23 from vapier/master

move interactive function init out of main loop

9 years agomove interactive function init out of main loop
Mike Frysinger [Wed, 3 Jun 2015 13:24:48 +0000 (21:24 +0800)]
move interactive function init out of main loop

These values don't change between runs, so there's no point in recreating
them on every instruction step.  Pull it out to the top to simplify.

9 years agoUse single, shared real-time counter
Andrew Waterman [Mon, 1 Jun 2015 01:28:53 +0000 (18:28 -0700)]
Use single, shared real-time counter

This required disentangling INSTRET/CYCLE from TIME.

9 years agoExecute exactly the # of insns passed to step()
Andrew Waterman [Mon, 1 Jun 2015 01:10:20 +0000 (18:10 -0700)]
Execute exactly the # of insns passed to step()

Previously, the value was treated as approximate.

9 years agoAdd rest of RV32C instructions
Andrew Waterman [Sun, 31 May 2015 23:38:24 +0000 (16:38 -0700)]
Add rest of RV32C instructions

9 years agoFix performance bug when CSR accesses are common
Andrew Waterman [Sun, 31 May 2015 22:47:11 +0000 (15:47 -0700)]
Fix performance bug when CSR accesses are common

processor_t::step was returning without doing much work, causing the HTIF
to be ticked far more often than necessary.

9 years agoFix c.slliw implementation
Andrew Waterman [Thu, 21 May 2015 18:55:52 +0000 (11:55 -0700)]
Fix c.slliw implementation

9 years agoNew RV64C proposal
Andrew Waterman [Thu, 21 May 2015 07:21:46 +0000 (00:21 -0700)]
New RV64C proposal

9 years agoTake interrupts as soon as interrupts are enabled
Andrew Waterman [Wed, 20 May 2015 20:49:01 +0000 (13:49 -0700)]
Take interrupts as soon as interrupts are enabled

Previously, if interrupts were enabled then disabled quickly enough,
no interrupt would ever be taken, resulting in deadlock.

9 years agoMerge pull request #20 from palmer-dabbelt/package
Andrew Waterman [Fri, 15 May 2015 00:35:18 +0000 (17:35 -0700)]
Merge pull request #20 from palmer-dabbelt/package

various fixes to make riscv-isa-sim installable

9 years agoFix VM, MIP encoding
Andrew Waterman [Mon, 11 May 2015 21:21:03 +0000 (14:21 -0700)]
Fix VM, MIP encoding

9 years agoInstall "disasm.h"
Palmer Dabbelt [Wed, 13 May 2015 22:36:57 +0000 (15:36 -0700)]
Install "disasm.h"

Something includes this somewhere, so I see no reason not to just
install it.

9 years agoautoreconf 126beb482ff6 ("Install pkg-config files for Spike")
Palmer Dabbelt [Wed, 13 May 2015 22:36:45 +0000 (15:36 -0700)]
autoreconf 126beb482ff6 ("Install pkg-config files for Spike")

9 years agoInstall pkg-config files for Spike
Palmer Dabbelt [Wed, 13 May 2015 21:54:24 +0000 (14:54 -0700)]
Install pkg-config files for Spike

The actual Spike package consists of a whole bunch of libraries.  This
installs a pkg-config file for each generated library, and a meta
pkg-config file (riscv-spike.pc) that links in all the other Spike
libraries except dummy_rocc (which I figure isn't interesting).

This allows me to package and install spike, and then use that package
to build an extension library (so I don't have to fork Spike for my
own ISA extension).

9 years agoChange the behavior of the DESTDIR make variable
Palmer Dabbelt [Wed, 13 May 2015 21:52:19 +0000 (14:52 -0700)]
Change the behavior of the DESTDIR make variable

DESTDIR is a common make idiom.  As per the GNU coding standards

  https://www.gnu.org/prep/standards/html_node/DESTDIR.html

  "DESTDIR is a variable prepended to each installed target file, like
  this:

  $(INSTALL_PROGRAM) foo $(DESTDIR)$(bindir)/foo
  $(INSTALL_DATA) libfoo.a $(DESTDIR)$(libdir)/libfoo.a

  The DESTDIR variable is specified by the user on the make command
  line as an absolute file name. For example:

  make DESTDIR=/tmp/stage install

  DESTDIR should be supported only in the install* and uninstall*
  targets, as those are the only targets where it is useful.

  If your installation step would normally install /usr/local/bin/foo
  and /usr/local/lib/libfoo.a, then an installation invoked as in the
  example above would install /tmp/stage/usr/local/bin/foo and
  /tmp/stage/usr/local/lib/libfoo.a instead."

The current Makefile.in uses DESTDIR, but has a slightly non-standard
behavior: the target install location doesn't include "$prefix".  This
breaks package managers, because stuff ends up getting installed to
the wrong location.

Unfortunately the only way I can think of to fix this involves
silently changing the behavior of DESTDIR.  Hopefully nobody is using
it...?

[port of 8a2088b59162fe16c16d26ddc1cfcaaaa8c4156f in riscv-fesvr]

9 years agoUpgrade to privileged architecture 1.7
Andrew Waterman [Sat, 9 May 2015 23:12:07 +0000 (16:12 -0700)]
Upgrade to privileged architecture 1.7

9 years agoFix commit log for CSR instructions
Andrew Waterman [Thu, 30 Apr 2015 21:31:25 +0000 (14:31 -0700)]
Fix commit log for CSR instructions

9 years agoFix I$ simulator hit count
Andrew Waterman [Sun, 26 Apr 2015 00:51:23 +0000 (17:51 -0700)]
Fix I$ simulator hit count

Also, improve performance by allowing the soft-ITLB to refill.

9 years agoMerge pull request #18 from wsong83/master
Andrew Waterman [Tue, 14 Apr 2015 22:46:30 +0000 (15:46 -0700)]
Merge pull request #18 from wsong83/master

fix cache line index offset in cachesim.cc

9 years agofix cache line index offset in cachesim.cc
Wei Song [Mon, 13 Apr 2015 15:35:13 +0000 (16:35 +0100)]
fix cache line index offset in cachesim.cc

9 years agoFix autoconf for RHEL6
Andrew Waterman [Thu, 9 Apr 2015 18:25:07 +0000 (11:25 -0700)]
Fix autoconf for RHEL6

9 years agoCheck for F extension when accessing FCSR
Andrew Waterman [Sat, 4 Apr 2015 22:17:13 +0000 (15:17 -0700)]
Check for F extension when accessing FCSR

9 years agoSupport setting ISA/subsets with --isa flag
Andrew Waterman [Sat, 4 Apr 2015 04:53:22 +0000 (21:53 -0700)]
Support setting ISA/subsets with --isa flag

Default is RV64IMAFDC.  Can do things like

  --isa=RV32 (which implies IMAFDC)
  --isa=IM (which implies RV64)
  --isa=RV64IMAFDXhwacha

9 years agoSimplify RV32 comparisons
Andrew Waterman [Thu, 2 Apr 2015 21:23:30 +0000 (14:23 -0700)]
Simplify RV32 comparisons

No need to eliminate the upper 32 bits of the 64-bit x-register, as all
RV32 instructions should sign-extend their results to 64 bits.

9 years agoAllow writing mstatus.fs even if FPU isn't present
Andrew Waterman [Tue, 31 Mar 2015 23:07:56 +0000 (16:07 -0700)]
Allow writing mstatus.fs even if FPU isn't present

This allows the OS to track FP state dirtiness.

9 years agoImplement RVC draft
Andrew Waterman [Tue, 31 Mar 2015 03:28:55 +0000 (20:28 -0700)]
Implement RVC draft

9 years agoSerialize counters without throwing C++ exceptions
Andrew Waterman [Thu, 26 Mar 2015 07:28:10 +0000 (00:28 -0700)]
Serialize counters without throwing C++ exceptions

Ideally, a similar mechanism will apply to target machine exceptions.

9 years agoNew virtual memory implementation (Sv39)
Andrew Waterman [Tue, 24 Mar 2015 22:29:43 +0000 (15:29 -0700)]
New virtual memory implementation (Sv39)

9 years agoUpdate state.pc on every instruction
Andrew Waterman [Thu, 26 Mar 2015 06:01:54 +0000 (23:01 -0700)]
Update state.pc on every instruction

This isn't a bug fix for Spike proper, but it makes it possible for
RoCC instructions to access the control thread's PC.

9 years agoFor misaligned fetch, set mepc = addr of branch/jump
Andrew Waterman [Sat, 21 Mar 2015 06:11:49 +0000 (23:11 -0700)]
For misaligned fetch, set mepc = addr of branch/jump

9 years agobugfix, mbadaddr should be writable
Yunsup Lee [Tue, 17 Mar 2015 11:33:17 +0000 (04:33 -0700)]
bugfix, mbadaddr should be writable

9 years agochange hwacha cause to follow risc-v cause
Yunsup Lee [Tue, 17 Mar 2015 11:32:53 +0000 (04:32 -0700)]
change hwacha cause to follow risc-v cause

9 years agoMerge [shm]call into ecall, [shm]ret into eret
Andrew Waterman [Tue, 17 Mar 2015 08:19:40 +0000 (01:19 -0700)]
Merge [shm]call into ecall, [shm]ret into eret

9 years agobugfix in raising accelerator interrupts
Yunsup Lee [Mon, 16 Mar 2015 09:01:26 +0000 (02:01 -0700)]
bugfix in raising accelerator interrupts

9 years agovxcptsave->vxcptevac,vxcptrestore->vxcpthold
Yunsup Lee [Mon, 16 Mar 2015 08:59:55 +0000 (01:59 -0700)]
vxcptsave->vxcptevac,vxcptrestore->vxcpthold

this change was necessary since the impl csr is gone.

9 years agoDelete all remaining junk on "make clean"
Andrew Waterman [Mon, 16 Mar 2015 03:41:24 +0000 (20:41 -0700)]
Delete all remaining junk on "make clean"

9 years agoDon't set dirty/referenced bits w/o permission
Andrew Waterman [Sat, 14 Mar 2015 09:05:15 +0000 (02:05 -0700)]
Don't set dirty/referenced bits w/o permission

9 years agoUse hcall instead of mcall
Andrew Waterman [Fri, 13 Mar 2015 06:06:40 +0000 (23:06 -0700)]
Use hcall instead of mcall

9 years agoImplement PTE referenced/dirty bits
Andrew Waterman [Fri, 13 Mar 2015 03:00:09 +0000 (20:00 -0700)]
Implement PTE referenced/dirty bits

9 years agoUpdate to new privileged spec
Andrew Waterman [Fri, 13 Mar 2015 00:32:43 +0000 (17:32 -0700)]
Update to new privileged spec

Sorry, everyone.

9 years agocorrectly parse extension for spike-disasm
Yunsup Lee [Thu, 12 Mar 2015 17:59:09 +0000 (10:59 -0700)]
correctly parse extension for spike-disasm

9 years agoSoftfloat fcvt.{w/l}.d now returns -MaxInt for -NaN
Andrew Waterman [Tue, 24 Feb 2015 02:35:14 +0000 (18:35 -0800)]
Softfloat fcvt.{w/l}.d now returns -MaxInt for -NaN

9 years agoSoftfloat fcvt.{w/h}.s now returns -MaxInt for -NaN
Andrew Waterman [Mon, 23 Feb 2015 03:27:37 +0000 (19:27 -0800)]
Softfloat fcvt.{w/h}.s now returns -MaxInt for -NaN

   Behavior now consistent with RISC-V user spec.

9 years agohandle Makefile dependences for precompiled headers correctly
Andrew Waterman [Sun, 15 Feb 2015 00:38:48 +0000 (16:38 -0800)]
handle Makefile dependences for precompiled headers correctly

9 years agoInstall config.h with other headers
Andrew Waterman [Tue, 10 Feb 2015 01:42:19 +0000 (17:42 -0800)]
Install config.h with other headers

9 years agoUse xlen, not xprlen, to refer to x-register width
Andrew Waterman [Sun, 8 Feb 2015 08:13:48 +0000 (00:13 -0800)]
Use xlen, not xprlen, to refer to x-register width

9 years agoproperly sign-extend instructions in spike-dasm
Andrew Waterman [Thu, 29 Jan 2015 20:49:30 +0000 (12:49 -0800)]
properly sign-extend instructions in spike-dasm

9 years agoFixed masking/casting logic in commit log printf.
Christopher Celio [Tue, 27 Jan 2015 08:32:57 +0000 (00:32 -0800)]
Fixed masking/casting logic in commit log printf.

9 years agoFix commit log
Andrew Waterman [Mon, 26 Jan 2015 22:01:39 +0000 (14:01 -0800)]
Fix commit log

I screwed up some stuff in a recent refactoring.

9 years agoDon't look for a simulator when cross compiling
Andrew Waterman [Sat, 10 Jan 2015 01:41:59 +0000 (17:41 -0800)]
Don't look for a simulator when cross compiling

9 years agoUse PATH_MAX instead of NAME_MAX
Andrew Waterman [Sat, 10 Jan 2015 01:41:43 +0000 (17:41 -0800)]
Use PATH_MAX instead of NAME_MAX

9 years agoFix bug where C compiler used instead of C++ for autoconf tests
Stephen Twigg [Fri, 9 Jan 2015 23:56:28 +0000 (15:56 -0800)]
Fix bug where C compiler used instead of C++ for autoconf tests

9 years agocanonicalize assembler pseudo-ops
Andrew Waterman [Mon, 5 Jan 2015 03:40:13 +0000 (19:40 -0800)]
canonicalize assembler pseudo-ops

9 years agoRename riscv-dis to spike-dasm
Andrew Waterman [Mon, 5 Jan 2015 01:31:11 +0000 (17:31 -0800)]
Rename riscv-dis to spike-dasm

This is a better name, since it looks for the string DASM(xxx), and it
fixes tab completion for the cross compiler.

9 years agoDisassemble jalr x0, x1, 0 as ret
Andrew Waterman [Mon, 5 Jan 2015 01:30:40 +0000 (17:30 -0800)]
Disassemble jalr x0, x1, 0 as ret

9 years agoRequire 4-byte instruction alignment until RVC is reimplemented
Andrew Waterman [Sat, 3 Jan 2015 07:17:44 +0000 (23:17 -0800)]
Require 4-byte instruction alignment until RVC is reimplemented

9 years agoOn misaligned fetch, set EPC to target, not branch itself
Andrew Waterman [Sat, 3 Jan 2015 01:29:05 +0000 (17:29 -0800)]
On misaligned fetch, set EPC to target, not branch itself

9 years agoReduce dependences on auto-generated code
Andrew Waterman [Fri, 2 Jan 2015 23:09:04 +0000 (15:09 -0800)]
Reduce dependences on auto-generated code

In particular, precompiled headers ideally won't depend on any.

9 years agoMerge branch 'netbsd-autoconf'
Palmer Dabbelt [Tue, 30 Dec 2014 21:24:51 +0000 (13:24 -0800)]
Merge branch 'netbsd-autoconf'

This contains some changes that fix the build on NetBSD.

9 years agoautoreconf 65ba70071d11cc19b3dc85c047c5fea6d4d7bc0d
Palmer Dabbelt [Mon, 29 Dec 2014 21:07:13 +0000 (13:07 -0800)]
autoreconf 65ba70071d11cc19b3dc85c047c5fea6d4d7bc0d

9 years agoAllow Spike to build on NetBSD
Reinoud Zandijk [Mon, 29 Dec 2014 21:06:09 +0000 (13:06 -0800)]
Allow Spike to build on NetBSD

Under NetBSD, dlopen() is in libc so no extra libs are needed.

9 years agoSupport building from within root directory
Andrew Waterman [Sat, 20 Dec 2014 21:37:14 +0000 (13:37 -0800)]
Support building from within root directory

9 years agoFix makefile race condition
Andrew Waterman [Sat, 20 Dec 2014 21:26:48 +0000 (13:26 -0800)]
Fix makefile race condition

9 years agoMerge pull request #11 from arunthomas/readme
Andrew Waterman [Fri, 12 Dec 2014 00:52:56 +0000 (16:52 -0800)]
Merge pull request #11 from arunthomas/readme

README: use gnu-toolchain

9 years agoREADME: use gnu-toolchain
Arun Thomas [Thu, 11 Dec 2014 16:20:22 +0000 (11:20 -0500)]
README: use gnu-toolchain

9 years agozero-extend 32b instructions for vxcptaux
Andrew Waterman [Fri, 5 Dec 2014 09:08:27 +0000 (01:08 -0800)]
zero-extend 32b instructions for vxcptaux

9 years agoSupport 2/4/6/8-byte instructions
Andrew Waterman [Fri, 5 Dec 2014 07:10:33 +0000 (23:10 -0800)]
Support 2/4/6/8-byte instructions

Most of the complexity is in instruction address translation, since
instructions may span page boundaries.

9 years agoSet badvaddr on instruction page faults
Andrew Waterman [Fri, 5 Dec 2014 07:08:01 +0000 (23:08 -0800)]
Set badvaddr on instruction page faults

This supports distinguishing the EPC (the address of the first byte of the
faulting instruction) from the address of the page fault (potentially some
bytes later).