linux_dsm_epyc7002/arch/arm64/kernel
Daniel Borkmann 85f68fe898 bpf, arm64: implement jiting of BPF_XADD
This work adds BPF_XADD for BPF_W/BPF_DW to the arm64 JIT and therefore
completes JITing of all BPF instructions, meaning we can thus also remove
the 'notyet' label and do not need to fall back to the interpreter when
BPF_XADD is used in a program!

This now also brings arm64 JIT in line with x86_64, s390x, ppc64, sparc64,
where all current eBPF features are supported.

BPF_W example from test_bpf:

  .u.insns_int = {
    BPF_ALU32_IMM(BPF_MOV, R0, 0x12),
    BPF_ST_MEM(BPF_W, R10, -40, 0x10),
    BPF_STX_XADD(BPF_W, R10, R0, -40),
    BPF_LDX_MEM(BPF_W, R0, R10, -40),
    BPF_EXIT_INSN(),
  },

  [...]
  00000020:  52800247  mov w7, #0x12 // #18
  00000024:  928004eb  mov x11, #0xffffffffffffffd8 // #-40
  00000028:  d280020a  mov x10, #0x10 // #16
  0000002c:  b82b6b2a  str w10, [x25,x11]
  // start of xadd mapping:
  00000030:  928004ea  mov x10, #0xffffffffffffffd8 // #-40
  00000034:  8b19014a  add x10, x10, x25
  00000038:  f9800151  prfm pstl1strm, [x10]
  0000003c:  885f7d4b  ldxr w11, [x10]
  00000040:  0b07016b  add w11, w11, w7
  00000044:  880b7d4b  stxr w11, w11, [x10]
  00000048:  35ffffab  cbnz w11, 0x0000003c
  // end of xadd mapping:
  [...]

BPF_DW example from test_bpf:

  .u.insns_int = {
    BPF_ALU32_IMM(BPF_MOV, R0, 0x12),
    BPF_ST_MEM(BPF_DW, R10, -40, 0x10),
    BPF_STX_XADD(BPF_DW, R10, R0, -40),
    BPF_LDX_MEM(BPF_DW, R0, R10, -40),
    BPF_EXIT_INSN(),
  },

  [...]
  00000020:  52800247  mov w7,  #0x12 // #18
  00000024:  928004eb  mov x11, #0xffffffffffffffd8 // #-40
  00000028:  d280020a  mov x10, #0x10 // #16
  0000002c:  f82b6b2a  str x10, [x25,x11]
  // start of xadd mapping:
  00000030:  928004ea  mov x10, #0xffffffffffffffd8 // #-40
  00000034:  8b19014a  add x10, x10, x25
  00000038:  f9800151  prfm pstl1strm, [x10]
  0000003c:  c85f7d4b  ldxr x11, [x10]
  00000040:  8b07016b  add x11, x11, x7
  00000044:  c80b7d4b  stxr w11, x11, [x10]
  00000048:  35ffffab  cbnz w11, 0x0000003c
  // end of xadd mapping:
  [...]

Tested on Cavium ThunderX ARMv8, test suite results after the patch:

  No JIT:   [ 3751.855362] test_bpf: Summary: 311 PASSED, 0 FAILED, [0/303 JIT'ed]
  With JIT: [ 3573.759527] test_bpf: Summary: 311 PASSED, 0 FAILED, [303/303 JIT'ed]

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-05-02 15:04:50 -04:00
..
probes arm64: kprobes: remove kprobe_exceptions_notify 2017-03-10 17:41:19 +00:00
vdso arm64: drop non-existing vdso-offsets.h from .gitignore 2017-03-30 19:29:30 +01:00
.gitignore arm64: Build infrastructure 2012-09-17 13:42:21 +01:00
acpi_numa.c arm64/numa: avoid inconsistent information to be printed 2016-09-09 14:59:08 +01:00
acpi_parking_protocol.c arm64: Use __pa_symbol for kernel symbols 2017-01-12 15:05:39 +00:00
acpi.c ACPI / osl: Remove acpi_get_table_with_size()/early_acpi_os_unmap_memory() users 2016-12-21 02:36:38 +01:00
alternative.c arm64: alternative: Add support for patching adrp instructions 2016-09-09 15:03:28 +01:00
arm64ksyms.c arm: kernel: Add SMC structure parameter 2017-02-03 18:46:33 +00:00
armv8_deprecated.c kprobes: move kprobe declarations to asm-generic/kprobes.h 2017-02-27 18:43:45 -08:00
asm-offsets.c arm: kernel: Add SMC structure parameter 2017-02-03 18:46:33 +00:00
cacheinfo.c arm64: cacheinfo: add support to override cache levels via device tree 2017-01-17 12:09:54 +00:00
cpu_errata.c arm64: Work around Falkor erratum 1003 2017-02-10 11:22:12 +00:00
cpu_ops.c arm64: apply __ro_after_init to some objects 2016-08-22 12:32:29 +01:00
cpu-reset.h arm64: Use __pa_symbol for kernel symbols 2017-01-12 15:05:39 +00:00
cpu-reset.S arm64: Add back cpu reset routines 2016-06-27 16:31:25 +01:00
cpufeature.c arm64: cpufeature: correctly handle MRS to XZR 2017-02-15 12:20:29 +00:00
cpuidle.c arm64: kernel: Update kerneldoc for cpu_suspend() rename 2017-03-10 18:06:24 +00:00
cpuinfo.c arm64: Advertise support for Rounding double multiply instructions 2017-01-12 17:19:06 +00:00
debug-monitors.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/task_stack.h> 2017-03-02 08:42:36 +01:00
efi-entry.S arm64: efi-entry.S: avoid open-coded adr_l 2017-01-17 17:41:14 +00:00
efi.c arm64: mm: replace 'block_mappings_allowed' with 'page_mappings_only' 2016-11-07 18:15:04 +00:00
entry32.S arm64: entry32: remove pointless register assignment 2015-07-10 16:47:13 +01:00
entry-fpsimd.S arm64: fpsimd: fix a typo in fpsimd_save_partial_state ENDPROC 2014-07-31 11:42:42 +01:00
entry-ftrace.S arm64: include asm/assembler.h in entry-ftrace.S 2017-02-15 11:34:25 +00:00
entry.S arm64: avoid returning from bad_mode 2017-01-19 15:38:22 +00:00
fpsimd.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/signal.h> 2017-03-02 08:42:29 +01:00
ftrace.c ftrace: Add return address pointer to ftrace_ret_stack 2016-08-24 12:15:14 +02:00
head.S arm64: head.S: Enable EL1 (host) access to SPE when entered at EL2 2017-02-09 18:31:25 +00:00
hibernate-asm.S arm64: Introduce raw_{d,i}cache_line_size 2016-09-09 15:03:29 +01:00
hibernate.c arm64: Use __pa_symbol for kernel symbols 2017-01-12 15:05:39 +00:00
hw_breakpoint.c cpu/hotplug: Cleanup state names 2016-12-25 10:47:44 +01:00
hyp-stub.S arm64: Add back cpu reset routines 2016-06-27 16:31:25 +01:00
image.h arm64 updates for 4.7: 2016-05-16 17:17:24 -07:00
insn.c bpf, arm64: implement jiting of BPF_XADD 2017-05-02 15:04:50 -04:00
io.c arm64: optimize memcpy_{from,to}io() and memset_io() 2014-11-06 17:25:27 +00:00
irq.c arm64: remove irq_count and do_softirq_own_stack() 2015-12-21 17:26:01 +00:00
jump_label.c jump_label: Rename JUMP_LABEL_{EN,DIS}ABLE to JUMP_LABEL_{JMP,NOP} 2015-08-03 11:34:12 +02:00
kaslr.c arm64: kaslr: Fix up the kernel image alignment 2017-03-22 15:43:11 +00:00
kgdb.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/task_stack.h> 2017-03-02 08:42:36 +01:00
kuser32.S arm64: Add __NR_* definitions for compat syscalls 2014-07-10 11:02:40 +01:00
machine_kexec.c arm64/kexec: Add pr_debug output 2016-06-27 16:31:26 +01:00
Makefile efi: arm64: Add vmlinux debug link to the Image binary 2017-02-03 15:22:37 +00:00
module-plts.c arm64: add support for module PLTs 2016-02-24 14:57:24 +00:00
module.c arm64: add support for kernel ASLR 2016-02-24 14:57:27 +00:00
module.lds arm64: add support for module PLTs 2016-02-24 14:57:24 +00:00
paravirt.c arm64: introduce CONFIG_PARAVIRT, PARAVIRT_TIME_ACCOUNTING and pv_time_ops 2015-12-21 14:40:54 +00:00
pci.c PCI/ACPI: Extend pci_mcfg_lookup() to return ECAM config accessors 2016-12-06 13:45:48 -06:00
perf_callchain.c perf core: Add a 'nr' field to perf_event_callchain_context 2016-05-16 23:11:51 -03:00
perf_event.c KVM: arm64: Fix the issues when guest PMCCFILTR is configured 2016-11-18 09:06:58 +00:00
perf_regs.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/task_stack.h> 2017-03-02 08:42:36 +01:00
process.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/task_stack.h> 2017-03-02 08:42:36 +01:00
psci.c arm64: Use __pa_symbol for kernel symbols 2017-01-12 15:05:39 +00:00
ptrace.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/task_stack.h> 2017-03-02 08:42:36 +01:00
relocate_kernel.S arm64: Introduce raw_{d,i}cache_line_size 2016-09-09 15:03:29 +01:00
return_address.c arm64: factor out current_stack_pointer 2016-11-11 18:25:43 +00:00
setup.c sched/headers: Prepare to move 'init_task' and 'init_thread_union' from <linux/sched.h> to <linux/sched/task.h> 2017-03-02 08:42:38 +01:00
signal32.c Replace <asm/uaccess.h> with <linux/uaccess.h> globally 2016-12-24 11:46:01 -08:00
signal.c arm64: Add uprobe support 2016-11-07 18:15:21 +00:00
sleep.S arm64: move sp_el0 and tpidr_el1 into cpu_suspend_ctx 2016-11-11 18:25:44 +00:00
smccc-call.S firmware: qcom: scm: Fix interrupted SCM calls 2017-02-03 18:46:33 +00:00
smp_spin_table.c arm64: Use __pa_symbol for kernel symbols 2017-01-12 15:05:39 +00:00
smp.c arm64: fix NULL dereference in have_cpu_die() 2017-03-30 17:14:32 +01:00
stacktrace.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/task_stack.h> 2017-03-02 08:42:36 +01:00
suspend.c arm64: move sp_el0 and tpidr_el1 into cpu_suspend_ctx 2016-11-11 18:25:44 +00:00
sys32.c arm64: fix implementation of mmap2 compat syscall 2015-03-19 10:43:51 +00:00
sys_compat.c sched/headers: Prepare for the reduction of <linux/sched.h>'s signal API dependency 2017-03-02 08:42:37 +01:00
sys.c arm64: compat: Check for AArch32 state 2016-04-20 12:22:42 +01:00
time.c arm64: ftrace: fix a stack tracer's output under function graph tracer 2015-12-21 17:26:02 +00:00
topology.c sched/headers: Prepare for new header dependencies before moving code to <linux/sched/topology.h> 2017-03-02 08:42:26 +01:00
trace-events-emulation.h arm64: Trace emulation of AArch32 legacy instructions 2014-11-20 16:35:02 +00:00
traps.c sched/headers: Prepare to remove the <linux/mm_types.h> dependency from <linux/sched.h> 2017-03-02 08:42:37 +01:00
vdso.c arm64: Use __pa_symbol for kernel symbols 2017-01-12 15:05:39 +00:00
vmlinux.lds.S arm64: Introduce uaccess_{disable,enable} functionality based on TTBR0_EL1 2016-11-21 18:48:53 +00:00