Tail calls allow JIT'ed eBPF programs to call into other JIT'ed eBPF
programs. This can be achieved either by:
(1) retaining the stack setup by the first eBPF program and having all
subsequent eBPF programs re-using it, or,
(2) by unwinding/tearing down the stack and having each eBPF program
deal with its own stack as it sees fit.
To ensure that this does not create loops, there is a limit to how many
tail calls can be done (currently 32). This requires the JIT'ed code to
maintain a count of the number of tail calls done so far.
Approach (1) is simple, but requires every eBPF program to have (almost)
the same prologue/epilogue, regardless of whether they need it. This is
inefficient for small eBPF programs which may not sometimes need a
prologue at all. As such, to minimize impact of tail call
implementation, we use approach (2) here which needs each eBPF program
in the chain to use its own prologue/epilogue. This is not ideal when
many tail calls are involved and when all the eBPF programs in the chain
have similar prologue/epilogue. However, the impact is restricted to
programs that do tail calls. Individual eBPF programs are not affected.
We maintain the tail call count in a fixed location on the stack and
updated tail call count values are passed in through this. The very
first eBPF program in a chain sets this up to 0 (the first 2
instructions). Subsequent tail calls skip the first two eBPF JIT
instructions to maintain the count. For programs that don't do tail
calls themselves, the first two instructions are NOPs.
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
PPC64 eBPF JIT compiler.
Enable with:
echo 1 > /proc/sys/net/core/bpf_jit_enable
or
echo 2 > /proc/sys/net/core/bpf_jit_enable
... to see the generated JIT code. This can further be processed with
tools/net/bpf_jit_disasm.
With CONFIG_TEST_BPF=m and 'modprobe test_bpf':
test_bpf: Summary: 305 PASSED, 0 FAILED, [297/297 JIT'ed]
... on both ppc64 BE and LE.
The details of the approach are documented through various comments in
the code.
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Break out classic BPF JIT specifics into a separate header in
preparation for eBPF JIT implementation. Note that ppc32 will still need
the classic BPF JIT.
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
1. Per the ISA, ADDIS actually uses RT, rather than RS. Though
the result is the same, make the usage clear.
2. The multiply instruction used is a 32-bit multiply. Rename PPC_MUL()
to PPC_MULW() to make the same clear.
3. PPC_STW[U] take the entire 16-bit immediate value and do not require
word-alignment, per the ISA. Change the macros to use IMM_L().
4. A few white-space cleanups to satisfy checkpatch.pl.
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Since we will be using the rotate immediate instructions for extended
BPF JIT, let's introduce macros for the same. And since the shift
immediate operations use the rotate immediate instructions, let's redo
those macros to use the newly introduced instructions.
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Similar to the LI32() optimization, if the value can be represented
in 32-bits, use LI32(). Also handle loading a few specific forms of
immediate values in an optimum manner.
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The existing LI32() macro can sometimes result in a sign-extended 32-bit
load that does not clear the top 32-bits properly. As an example,
loading 0x7fffffff results in the register containing
0xffffffff7fffffff. While this does not impact classic BPF JIT
implementation (since that only uses the lower word for all operations),
we would like to share this macro between classic BPF JIT and extended
BPF JIT, wherein the entire 64-bit value in the register matters. Fix
this by first doing a shifted LI followed by ORI.
An additional optimization is with loading values between -32768 to -1,
where we now only need a single LI.
The new implementation now generates the same or less number of
instructions.
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Use helpers from the asm-compat.h to wrap up assembly mnemonics
Signed-off-by: Denis Kirjanov <kda@linux-powerpc.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit b6069a9570 (filter: add MOD operation) added generic
support for modulus operation in BPF.
This patch brings JIT support for PPC64
Signed-off-by: Vladimir Murzin <murzin.v@gmail.com>
Acked-by: Matt Evans <matt@ozlabs.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
This enables the Berkeley Packet Filter JIT compiler
for the PowerPC running in 64bit Little Endian.
Signed-off-by: Philippe Bergheaud <felix@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
This patch is a follow-up for patch "filter: add XOR instruction for use
with X/K" that implements BPF PowerPC JIT parts for the BPF XOR operation.
Signed-off-by: Daniel Borkmann <daniel.borkmann@tik.ee.ethz.ch>
Cc: Matt Evans <matt@ozlabs.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: Matt Evans <matt@ozlabs.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now have ___PPC_RA/B/S/T we can use it in some places. These are
places where we can't use the existing defines which will soon enforce
R0-R31 usage.
The macros being changed here are being used in inline asm, which
can't convert to enforce the R0-R31 usage.
bpf_jit uses a mix of both generated and non-generated with the same
code, so just convert all these to use the ___PPC_R versions which
won't enforce R usage later.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Now the helper function from filter.c for negative offsets is exported,
it can be used it in the jit to handle negative offsets.
First modify the asm load helper functions to handle:
- know positive offsets
- know negative offsets
- any offset
then the compiler can be modified to explicitly use these helper
when appropriate.
This fixes the case of a negative X register and allows to lift
the restriction that bpf programs with negative offsets can't
be jited.
Tested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Jan Seiffert <kaffeemonster@googlemail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
An implementation of a code generator for BPF programs to speed up packet
filtering on PPC64, inspired by Eric Dumazet's x86-64 version.
Filter code is generated as an ABI-compliant function in module_alloc()'d mem
with stackframe & prologue/epilogue generated if required (simple filters don't
need anything more than an li/blr). The filter's local variables, M[], live in
registers. Supports all BPF opcodes, although "complicated" loads from negative
packet offsets (e.g. SKF_LL_OFF) are not yet supported.
There are a couple of further optimisations left for future work; many-pass
assembly with branch-reach reduction and a register allocator to push M[]
variables into volatile registers would improve the code quality further.
This currently supports big-endian 64-bit PowerPC only (but is fairly simple
to port to PPC32 or LE!).
Enabled in the same way as x86-64:
echo 1 > /proc/sys/net/core/bpf_jit_enable
Or, enabled with extra debug output:
echo 2 > /proc/sys/net/core/bpf_jit_enable
Signed-off-by: Matt Evans <matt@ozlabs.org>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Naveen N. Rao