Change 228019 by bz@bz_zenith on 2013/04/23 13:55:30
Add kernel side support for large TLB on BERI/CHERI.
Modelled similar to NLM
MFC after: 3 days
Sponsored by: DAPRA/AFRL
Change 221534 by rwatson@rwatson_zenith_cl_cam_ac_uk on 2013/01/27 16:05:30
FreeBSD/mips stores page-table entries in a near-identical format
to MIPS TLB entries -- only it overrides certain "reserved" bits
in the MIPS-defined EntryLo register to hold software-defined bits
(swbits) to avoid significantly increasing the page table memory
footprint. On n32 and n64, these bits were (a) colliding with
MIPS64r2 physical memory extensions and (b) being improperly
cleared.
Attempt to fix both of these problems by pushing swbits further
along 64-bit EntryLo registers into the reserved space, and
improving consistency between C-based and assembly-based clearing
of swbits -- in particular, to use the same definition. This
should stop swbits from leaking into TLB entries -- while ignored
by most current MIPS hardware, this would cause a problem with
(much) larger physical memory sizes, and also leads to confusing
hardware-level tracing as physical addresses contain unexpected
(and inconsistent) higher bits.
Discussed with: imp, jmallett
MFC after: 3 days
Sponsored by: DARPA/AFRL
* the mips74k cores only need EHB (which is 'sll $0, $0, 3')
here; NOPs don't actually work.
* add EHB as the last NOP for the default barriers/hazards;
that is "better" behaviour and should work on a wider
variety of processors.
This allows the existing (icky) TLB code to work, allowing
the AR9344 SoC (mips74k) to actually get through kernel startup.
Tested:
* AR9344 SoC - (mips74k)
* AR9331 SoC - (mips24k)
TODO:
* test on mips4k CPUs, just to be sure.
* document that sll $0, $0, 3 is actually "EHB" and that it
falls back to being a NOP for pre-mips32r1.
* mips24k has an errata that we currently don't correctly explicitly
state - ie, that after DERET/ERET, the only valid instruction is
a NOP.
Reviewed by: imp@
Approved by: re@ (gjb)
sf_buf_alloc()/sf_buf_free() inlines, to save two calls to an absolutely
empty functions.
Reviewed by: alc, kib, scottl
Sponsored by: Nginx, Inc.
Sponsored by: Netflix
They originated in the original Octeon port. They weren't present, as
far as I can tell, on the projects/mips branch until after this
point. They were in the original Octeon port in code picked up from
the vendor, who I've been able to find out trolling old email put them
there to get around an SMP problem that most likely was fixed in other
ways.
NetBSD and Linux don't have these, except for some specific uses of
SYNC on the alchemy parts (which we don't support, but even if we did
it is only a specific case and would be specifically coded
anyway). This is true of the current Linux code, as well as one old
version I polled.
I looked back at the old R12000, R8000, R6000, R4000, R4400 errata
that I have, and could find no mention of SYNC needing NOPs for
silicon bugs (although plenty of other cases where NOPs and other
contortions were needed).
An Google search turned up no old mailing list discussions on this on
Linux, NetBSD or FreeBSD (except the disussion that kicked off these
studies).
I've test booted this on my Octeon Plus eval board and survived a
buildworld. Adrian Chadd reports that this patch has no ill effects on
the Ahteros platforms he tested it on.
I conclude it is safe to just remove the NOPs. But added
__MIPS_PLATFORM_SYNC_NOPS as a failsafe in case we find some platform
where these are, in fact, required.
Reviewed by: adrian@
Issues were noted by Bruce Evans and are present on all architectures.
On i386, a counter fetch should use atomic read of 64bit value,
otherwise carry from the increment on other CPU could be lost for the
given fetch, making error of 2^32. If 64bit read (cmpxchg8b) is not
available on the machine, it cannot be SMP and it is enough to disable
preemption around read to avoid the split read.
On x86 the counter increment is not atomic on purpose, which makes it
possible for the store of the incremented result to override just
zeroed per-cpu slot. The effect would be a counter going off by
arbitrary value after zeroing. Perform the counter zeroing on the
same processor which does the increments, making the operations
mutually exclusive. On i386, same as for the fetching, if the
cmpxchg8b is not available, machine is not SMP and we disable
preemption for zeroing.
PowerPC64 is treated the same as amd64.
For other architectures, the changes made to allow the compilation to
succeed, without fixing the issues with zeroing or fetching. It
should be possible to handle them by using the 64bit loads and stores
atomic WRT preemption (assuming the architectures also converted from
using critical sections to proper asm). If architecture does not
provide the facility, using global (spin) mutex would be non-optimal
but working solution.
Noted by: bde
Sponsored by: The FreeBSD Foundation
order to match the MAXCPU concept. The change should also be useful
for consolidation and consistency.
Sponsored by: EMC / Isilon storage division
Obtained from: jeff
Reviewed by: alc
Remove #define to get kludges that asm.h used to define
Move clever macros to access assembler instructions to trap.c
Remove __ASSEMBLER__ ifdefs in regdef.h: they aren't needed anymore.
expand the %sccs.include.redist.c% directive with the standard
3-clause license, and add $FreeBSD$ to keep the commit script happy.
# This may break some mips stuff, which will be fixed in the next commit.
in the pcb. setjmp/longjmp in the kernel also used these values, so
continue to use them although their use isn't technically the pcb
register array (matching is all that's important for setjmp/longjmp in
the kernel). Finally, eliminate the old register names from regnum.h.
This is a lexical change only. The non-debug .o files have the same md5.
Partially implement generic_bs_*_8() for MIPS platforms.
This is known to work with TARGET_ARCH=mips64 with FreeBSD/BERI.
Assuming that other definitions in cpufunc.h are correct it will
work on non-o64 ABI systems except sibyte. On sibyte and o32 systems
generic_bs_*_8() will remain panic() implementations.
Sponsored by: DARPA, AFRL
Reviewed by: imp, jmallett (older versions)
to unique values.
There's some confusion about what the n32 assembler API really is
(since on page 9 of the spec they say that t0-t3 don't exist, then
turn around on page 22 and say that t4-t7 don't exist), and this
doesn't touch that.
NetBSD's version of this file follows the convention I used here, and
is likely to be correct.
This should fix gdb/ptrace.
Having MIPS_MAX_TLB_ENTRIES defined to 128 is misleading, since it used
to be 64 in older releases of MIPS architecture (where it could be read
from Config1) and can be much more than 128 for the newer processors.
For now, move the definition to the only file using it (mips/mips/tlb.c)
and define MIPS_MAX_TLB_ENTRIES depending on the MIPS cpu defined. Also
add few checks so that we do not write beyond the end of the tlb_state
array.
This fixes a kernel data corruption seen in Netlogic XLP, which was casued
by tlb_save() writing beyond the end of tlb_state array when breaking into
debugger.
Introduce counter(9) API, that implements fast and raceless counters,
provided (but not limited to) for gathering of statistical data.
See http://lists.freebsd.org/pipermail/freebsd-arch/2013-April/014204.html
for more details.
In collaboration with: kib
Reviewed by: luigi
Tested by: ae, ray
Sponsored by: Nginx, Inc.
rather than a constant so that VM_KMEM_SIZE_MAX will scale automatically
with the kernel address space size. This is particularly important for
MIPS because the same definition is used by both 32- and 64-bit kernels.
Tested by: jchandra
Provided a bus_space implementation for FDT, modelled on
bus_space_generic, but with a local version of the map address
routine that does a P->V translation, as is the case with NLM's
similar routine for XLP. It's not clear to me that this is the
right solution -- possibly this belongs in simplebus -- however,
it is sufficient to get the DE4 LED driver working.
Sponsored by: DARPA, AFRL
Add code so that the BERI boot process can ask the kernel linker for
DTB blobs that may have been left for it by the boot loader, as done
on PowerPC and ARM. This will require both a more mature boot
loader, and more mature boot loader argument passing mechanism,
than currently supported on BERI.
Sponsored by: DARPA, AFRL
page. Therefore, it is really inappropriate for use by the function
uma_small_alloc(). The effect of using it was that every page was zeroed
at least once and possibly twice if M_ZERO was passed as a "wait" flag.
instruction loads/stores at its will.
The macro __compiler_membar() is currently supported for both gcc and
clang, but kernel compilation will fail otherwise.
Reviewed by: bde, kib
Discussed with: dim, theraven
MFC after: 2 weeks
advantages. First, PV entries are roughly half the size. Second, this
allocator doesn't access the paging queues, and thus it will allow for the
removal of the page queues lock from this pmap.
Fix a rather serious bug in pmap_remove_write(). After removing write
access from the specified page's first mapping, pmap_remove_write() then
used the wrong "next" pointer. Consequently, the page's second, third,
etc. mappings were not write protected.
Tested by: jchandra
Reduce the size of a PV entry by eliminating pv_ptem. There is no need
to store a pointer to the page table page in the PV entry because it is
easily computed during the walk down the page table.
Eliminate the ptphint from the pmap. Long, long ago, page table pages
belonged to a vm object, and we would look up page table pages based
upon their offset within this vm object. In those days, this hint may
have had tangible benefits.
Tested by: jchandra
This is required for ARM EABI. Section 7.1.1 of the Procedure Call for the
ARM Architecture (AAPCS) defines wchar_t as either an unsigned int or an
unsigned short with the former preferred.
Because of this requirement we need to move the definition of __wchar_t to
a machine dependent header. It also cleans up the macros defining the limits
of wchar_t by defining __WCHAR_MIN and __WCHAR_MAX in the same machine
dependent header then using them to define WCHAR_MIN and WCHAR_MAX
respectively.
Discussed with: bde
usermode, using shared page. The structures and functions have vdso
prefix, to indicate the intended location of the code in some future.
The versioned per-algorithm data is exported in the format of struct
vdso_timehands, which mostly repeats the content of in-kernel struct
timehands. Usermode reading of the structure can be lockless.
Compatibility export for 32bit processes on 64bit host is also
provided. Kernel also provides usermode with indication about
currently used timecounter, so that libc can fall back to syscall if
configured timecounter is unknown to usermode code.
The shared data updates are initiated both from the tc_windup(), where
a fast task is queued to do the update, and from sysctl handlers which
change timecounter. A manual override switch
kern.timecounter.fast_gettime allows to turn off the mechanism.
Only x86 architectures export the real algorithm data, and there, only
for tsc timecounter. HPET counters page could be exported as well, but
I prefer to not further glue the kernel and libc ABI there until
proper vdso-based solution is developed.
Minimal stubs neccessary for non-x86 architectures to still compile
are provided.
Discussed with: bde
Reviewed by: jhb
Tested by: flo
MFC after: 1 month
layer, but it is read directly by the MI VM layer. This change introduces
pmap_page_is_write_mapped() in order to completely encapsulate all direct
access to PGA_WRITEABLE in the pmap layer.
Aesthetics aside, I am making this change because amd64 will likely begin
using an alternative method to track write mappings, and having
pmap_page_is_write_mapped() in place allows me to make such a change
without further modification to the MI VM layer.
As an added bonus, tidy up some nearby comments concerning page flags.
Reviewed by: kib
MFC after: 6 weeks
in_cksum.h required ip.h to be included for struct ip. To be
able to use some general checksum functions like in_addword()
in a non-IPv4 context, limit the (also exported to user space)
IPv4 specific functions to the times, when the ip.h header is
present and IPVERSION is defined (to 4).
We should consider more general checksum (updating) functions
to also allow easier incremental checksum updates in the L3/4
stack and firewalls, as well as ponder further requirements by
certain NIC drivers needing slightly different pseudo values
in offloading cases. Thinking in terms of a better "library".
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
This makes our naming scheme more closely match other systems and the
expectations of much third-party software. MIPS builds which are little-endian
should require and exhibit no changes. Big-endian TARGET_ARCHes must be
changed:
From: To:
mipseb mips
mipsn32eb mipsn32
mips64eb mips64
An entry has been added to UPDATING and some foot-shooting protection (complete
with warnings which should become errors in the near future) to the top-level
base system Makefile.
New kernel events can be added at various location for sampling or counting.
This will for example allow easy system profiling whatever the processor is
with known tools like pmcstat(8).
Simultaneous usage of software PMC and hardware PMC is possible, for example
looking at the lock acquire failure, page fault while sampling on
instructions.
Sponsored by: NETASQ
MFC after: 1 month
