The change makes the user and kernel address spaces on i386
independent, giving each almost the full 4G of usable virtual addresses
except for one PDE at top used for trampoline and per-CPU trampoline
stacks, and system structures that must be always mapped, namely IDT,
GDT, common TSS and LDT, and process-private TSS and LDT if allocated.
By using 1:1 mapping for the kernel text and data, it appeared
possible to eliminate assembler part of the locore.S which bootstraps
initial page table and KPTmap. The code is rewritten in C and moved
into the pmap_cold(). The comment in vmparam.h explains the KVA
layout.
There is no PCID mechanism available in protected mode, so each
kernel/user switch forth and back completely flushes the TLB, except
for the trampoline PTD region. The TLB invalidations for userspace
becomes trivial, because IPI handlers switch page tables. On the other
hand, context switches no longer need to reload %cr3.
copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for
new copyout(9) is compatibility with wiring user buffers around sysctl
handlers. This explains two kind of locks for copyout ptes and
accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow
path, is only tried after the 'fast path' failed, which temporary
changes mapping to the userspace and copies the data to/from small
per-cpu buffer in the trampoline. If a page fault occurs during the
copy, it is short-circuit by exception.s to not even reach C code.
The change was motivated by the need to implement the Meltdown
mitigation, but instead of KPTI the full split is done. The i386
architecture already shows the sizing problems, in particular, it is
impossible to link clang and lld with debugging. I expect that the
issues due to the virtual address space limits would only exaggerate
and the split gives more liveness to the platform.
Tested by: pho
Discussed with: bde
Sponsored by: The FreeBSD Foundation
MFC after: 1 month
Differential revision: https://reviews.freebsd.org/D14633
Even though gdb and kgdb may not be removed for 12.0 on some architectures,
the notice is unconditional as these tools will likely be removed at some
point in the future when adequate replacements are available (gdb in ports
or lldb in base).
Reviewed by: emaste
MFC after: 3 days
Differential Revision: https://reviews.freebsd.org/D11477
Modify trapframe decoding to properly analyze trapframe.
Provide method for fixup_pc. It happens, that in some kernel
functions, the GDB stack frame decoder cannot determine both
func name and frame size. This is because these functions
either contain invalid instruction, or their format does
not match standard schema. Detect that scenarios and move
PC accordingly to jump into known function schema, which
GDB is able to parse.
Obtained from: Semihalf
Sponsored by: Juniper Networks
Reviewed by: kib, zbb
Differential Revision: https://reviews.freebsd.org/D5976
These are no longer needed after the recent 'beforebuild: depend' changes
and hooking DIRDEPS_BUILD into a subset of FAST_DEPEND which supports
skipping 'make depend'.
Sponsored by: EMC / Isilon Storage Division
Unfortunately filemon/meta mode tracks all indirect dependencies here
since ld(1) is reading libelf when linking in libkvm. Churn would be
reduced if this was able to be limited to direct dependencies.
Sponsored by: EMC / Isilon Storage Division
- Add a kvaddr_type to represent kernel virtual addresses instead of
unsigned long.
- Add a struct kvm_nlist which is a stripped down version of struct nlist
that uses kvaddr_t for n_value.
- Add a kvm_native() routine that returns true if an open kvm descriptor
is for a native kernel and memory image.
- Add a kvm_open2() function similar to kvm_openfiles(). It drops the
unused 'swapfile' argument and adds a new function pointer argument for
a symbol resolving function. Native kernels still use _fdnlist() from
libc to resolve symbols if a resolver function is not supplied, but cross
kernels require a resolver.
- Add a kvm_nlist2() function similar to kvm_nlist() except that it uses
struct kvm_nlist instead of struct nlist.
- Add a kvm_read2() function similar to kvm_read() except that it uses
kvaddr_t instead of unsigned long for the kernel virtual address.
- Add a new kvm_arch switch of routines needed by a vmcore backend.
Each backend is responsible for implementing kvm_read2() for a given
vmcore format.
- Use libelf to read headers from ELF kernels and cores (except for
powerpc cores).
- Add internal helper routines for the common page offset hash table used
by the minidump backends.
- Port all of the existing kvm backends to implement a kvm_arch switch and
to be cross-friendly by using private constants instead of ones that
vary by platform (e.g. PAGE_SIZE). Static assertions are present when
a given backend is compiled natively to ensure the private constants
match the real ones.
- Enable all of the existing vmcore backends on all platforms. This means
that libkvm on any platform should be able to perform KVA translation
and read data from a vmcore of any platform.
Tested on: amd64, i386, sparc64 (marius)
Differential Revision: https://reviews.freebsd.org/D3341
The kernel dump does not store these values on the stack.
Use PCB structure to resolve PC and LR properly.
Submitted by: Wojciech Macek <wma@semihalf.com>
Reviewed by: jhb, kib
Obtained from: Semihalf
Sponsored by: Juniper Networks Inc.
Differential Revision: https://reviews.freebsd.org/D4013
out of 'struct pcb' and into a variable-sized region after the
structure. The kgdb code currently only reads the pcb. It does not
read in the FPU save area but instead passes stack garbage as the
FPU's saved context. Fixing this would mean determining the proper
size of the area and fetching it. However, this state is not saved
for running CPUs in stoppcbs[], so the callback would also have to
know to ignore those pcbs. Instead, just remove the call since it is
of limited usefulness. It results in kgdb reporting the state of the
FPU/SIMD registers in userland, not their current values in the kernel.
In particular, it does not report the correct state for any code in
the kernel which does use the FPU and would report incorrect values
in that case.
Reviewed by: kib
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D3743
macros on amd64 and i386. Move the definition to machine/param.h.
kgdb defines INKERNEL() too, the conflict is resolved by renaming kgdb
version to PINKERNEL().
On i386, correct the lowest kernel address. After the shared page was
introduced, USRSTACK no longer points to the last user address + 1 [*]
Submitted by: Oliver Pinter [*]
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Off by default, build behaves normally.
WITH_META_MODE we get auto objdir creation, the ability to
start build from anywhere in the tree.
Still need to add real targets under targets/ to build packages.
Differential Revision: D2796
Reviewed by: brooks imp
- Eliminate unused irqframe
- Eliminate unused saframe
- Instead of splitting r4-sp storage between the stack and switchframe,
just put all the registers in switchframe and eliminate the un_32 struct.
Submitted by: Svatopluk Kraus <onwahe@gmail.com>,
Michal Meloun <meloun@miracle.cz>
comparing it with NOCPU, which became -1 recently. While here, avoid
using it for address calculations if it is negative.
Reviewed by: jhb, adrian
MFC after: 1 week
1. 50+% of NO_PIE use is fixed by adding -fPIC to INTERNALLIB and other
build-only utility libraries.
2. Another 40% is fixed by generating _pic.a variants of various libraries.
3. Some of the NO_PIE use is a bit absurd as it is disabling PIE (and ASLR)
where it never would work anyhow, such as csu or loader. This suggests
there may be better ways of adding support to the tree. Many of these
cases can be fixed such that -fPIE will work but there is really no
reason to have it in those cases.
4. Some of the uses are working around hacks done to some Makefiles that are
really building libraries but have been using bsd.prog.mk because the code
is cleaner. Had they been using bsd.lib.mk then NO_PIE would not have
been needed.
We likely do want to enable PIE by default (opt-out) for non-tree consumers
(such as ports). For in-tree though we probably want to only enable PIE
(opt-in) for common attack targets such as remote service daemons and setuid
utilities. This is also a great performance compromise since ASLR is expected
to reduce performance. As such it does not make sense to enable it in all
utilities such as ls(1) that have little benefit to having it enabled.
Reported by: kib
Make sure everything linking to a privatelib and/or an internallib does it directly
from the OBJDIR rather than DESTDIR.
Add src.libnames.mk so bsd.libnames.mk is not polluted by libraries not existsing
in final installation
Introduce the LD* variable which is what ld(1) is expecting (via LDADD) to link to
internal/privatelib
Directly link to the .so in case of private library to avoid having to complexify
LDFLAGS.
Phabric: https://phabric.freebsd.org/D553
Reviewed by: imp, emaste
variants. This allows usable file system images (i.e. those with both a
shell and an editor) to be created with only one copy of the curses library.
Exp-run: antoine
PR: 189842
Discussed with: bapt
Sponsored by: DARPA, AFRL
statically linked into consumers (GDB and variants) in the base
system, and the shared library is no longer installed.
That also allows ports to be able to use a modern version of readline
PR: 162948
Reviewed by: emaste
This includes:
o All directories named *ia64*
o All files named *ia64*
o All ia64-specific code guarded by __ia64__
o All ia64-specific makefile logic
o Mention of ia64 in comments and documentation
This excludes:
o Everything under contrib/
o Everything under crypto/
o sys/xen/interface
o sys/sys/elf_common.h
Discussed at: BSDcan
This is currently an opt-in build flag. Once ASLR support is ready and stable
it should changed to opt-out and be enabled by default along with ASLR.
Each application Makefile uses opt-out to ensure that ASLR will be enabled by
default in new directories when the system is compiled with PIE/ASLR. [2]
Mark known build failures as NO_PIE for now.
The only known runtime failure was rtld.
[1] http://www.bsdcan.org/2014/schedule/events/452.en.html
Submitted by: Shawn Webb <lattera@gmail.com>
Discussed between: des@ and Shawn Webb [2]
standard core target by declaring coreops_suppress_target with
initializer. This is also happening for non-cross kgdb, by
virtue of having fbsd-threads.c in libgdb and having it do the
exact same thing. Since fbsd-threads.c is not included in in
libgdb when building a cross debugger, we ended up with more
than 1 core file targets (the standard gdb core file target and
kgdb's libkvm based core file target) and this behaves the same
as not having a core target at all.
When looking up the absolute path for a kld, call find_kld_path() first.
This enables locating the module in a different directory than the one
stored in kernel memory.
With this change, kgdb can now be run on a kernel & vmcore whose associated
modules are located in the same directory as the kernel. This makes
independent triaging of problems much easier.
This change also does not break the normal kgdb use case where no arguments
are specified; in that case kgdb loads the running kernel and its modules.
Reviewed by: adrian
Approved by: ken (mentor)
Sponsored by: Spectra Logic
MFC after: 1 month
* document the kgdb -b flag
* better verify what's valid with -b
* add more comprehensive command line help
PR: kern/175743
Submitted by: Christoph Mallon <christoph.mallon@gmx.de>
on a core, when the core was stopped, by calling kgdb_trgt_core_pcb().
This has 2 advantages:
1. We don't need to include a machine-specific header anymore and as
such kthr.c is truly machine independent. This allows the code to
be used in a cross-debugger.
2. We don't need to lookup stoppcbs in generic code when it's an
inherently target-spicific symbol. It does not exist for ia64.
Implement kgdb_trgt_core_pcb() for all architectures, except ia64, by
calling a common function called kgdb_trgt_stop_pcb(). This function
differs from kgdb_trgt_core_pcb() in that it gets the size of the PCB
structure as an argument and as such remains machine independent.
On ia64 the PCB for stopped cores is in the PCPU structure itself.
This for better scaling. The implementation of kgdb_trgt_core_pcb()
for ia64 uses the cpuid_to_pcpu[] array to to obtain the address of
the PCB structure.
This allows a remote session to be specified with '-r' as well as a
non-default baudrate setting using '-b'.
TODO: add to the kgdb manpage.
MFC after: 2 weeks
