binutils-gdb/ld/testsuite/ld-mips-elf/micromips-branch-absolute.d

20 lines
674 B
D
Raw Normal View History

BFD: Let targets handle relocations against absolute symbols Fix a generic BFD issue with relocations against absolute symbols, which are installed without using any individual relocation handler provided by the backend. This causes any absolute section's addend to be lost on REL targets such as o32 MIPS, and also relocation-specific calculation adjustments are not made. As an example assembling this program: $ cat test.s .text foo: b bar b baz .set bar, 0x1234 $ as -EB -32 -o test-o32.o test.s $ as -EB -n32 -o test-n32.o test.s produces this binary code: $ objdump -dr test-o32.o test-n32.o test-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop test-n32.o: file format elf32-ntradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS*+0x1230 4: 00000000 nop 8: 10000000 b c <foo+0xc> 8: R_MIPS_PC16 baz-0x4 c: 00000000 nop $ where it is clearly visible in `test-o32.o', which uses REL relocations, that the absolute section's addend equivalent to the value of `bar' -- a reference to which cannot be fully resolved at the assembly time, because the reference is PC-relative -- has been lost, as has been the relocation-specific adjustment of -4, required to take into account the PC+4-relative calculation made by hardware with branches and seen in the external symbol reference to `baz' as the `ffff' addend encoded in the instruction word. In `test-n32.o', which uses RELA relocations, the absolute section's addend has been correctly retained. Give precedence then in `bfd_perform_relocation' and `bfd_install_relocation' to any individual relocation handler the backend selected may have provided, while still resorting to the generic calculation otherwise. This retains the semantics which we've had since forever or before the beginning of our repository history, and is at the very least compatible with `bfd_elf_generic_reloc' being used as the handler. Retain the `bfd_is_und_section' check unchanged at the beginning of `bfd_perform_relocation' since this does not affect the semantics of the function. The check returns the same `bfd_reloc_undefined' code the check for a null `howto' does, so swapping the two does not matter. Also the check is is mutually exclusive with the `bfd_is_abs_section' check, since a section cannot be absolute and undefined both at once, so swapping the two does not matter either. With this change applied the program quoted above now has the in-place addend correctly calculated and installed in the field being relocated: $ objdump -dr fixed-o32.o fixed-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 1000048c b 1234 <bar> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop $ Add a set of MIPS tests to cover the relevant cases, including absolute symbols with addends, and verifying that PC-relative relocations against symbols concerned resolve to the same value in the final link regardless of whether the REL or the RELA relocation form is used. Exclude linker tests though which would overflow the in-place addend on REL targets and use them as dump patterns for RELA targets only. bfd/ * reloc.c (bfd_perform_relocation): Try the `howto' handler first with relocations against absolute symbols. (bfd_install_relocation): Likewise. gas/ * testsuite/gas/mips/mips16-branch-absolute.d: Update patterns. * testsuite/gas/mips/branch-absolute.d: New test. * testsuite/gas/mips/branch-absolute-n32.d: New test. * testsuite/gas/mips/branch-absolute-n64.d: New test. * testsuite/gas/mips/branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/branch-absolute.s: New test source. * testsuite/gas/mips/branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips16-branch-absolute-addend.s: New test source. * testsuite/gas/mips/micromips-branch-absolute.s: New test source. * testsuite/gas/mips/micromips-branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips.exp: Run the new tests. ld/ * testsuite/ld-mips-elf/branch-absolute.d: New test. * testsuite/ld-mips-elf/branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/mips-elf.exp: Run the new tests, except from `branch-absolute-addend' and `micromips-branch-absolute-addend', referred indirectly only.
2016-07-12 08:30:01 +08:00
#objdump: -dr --prefix-addresses --show-raw-insn
#name: microMIPS link branch to absolute expression
#source: ../../../gas/testsuite/gas/mips/micromips-branch-absolute.s
#as: -EB -32
#ld: -EB -Ttext 0 -e foo
.*: +file format .*mips.*
Disassembly of section \.text:
\.\.\.
MIPS/GAS: Implement microMIPS branch/jump compaction Convert microMIPS branches and jumps whose delay slot would be filled by a generated NOP instruction to the corresponding compact form where one exists, in a manner similar to MIPS16 JR->JRC and JALR->JALRC swap. Do so even where the transformation switches from a 16-bit to a 32-bit branch encoding for no benefit in code size reduction, as this is still advantageous. This is because a branch/NOP pair takes 2 pipeline slots or a 2-cycle completion latency except in superscalar implementations. Whereas a compact branch may or may not stall on its target fetch, so it will at most have a 2-cycle completion latency and may have only 1 even in scalar implementations, and in superscalar implementations it is expected to have no worse latency as a branch/NOP pair has. Also it won't stall and therefore take the extra latency cycle in the not-taken case. Technically this is the same as MIPS16 compaction: for the qualifying instruction encodings the APPEND_ADD_COMPACT machine code generation method is selected where APPEND_ADD_WITH_NOP otherwise would and tells the code generator in `append_insn' to convert the regular form of an instruction to its corresponding compact form. For this the opcode is tweaked as necessary and the microMIPS opcode table is scanned for the matching updated instruction. A non-$0 `rt' operand to BEQ and BNE instructions is moved to the `rs' operand field of BEQZC and BNEZC encodings as required. Unlike with MIPS16 compaction however we need to handle out-of-distance branch relaxation as well. We do this by deferring the generation of any delay-slot NOP required to relaxation made in `md_convert_frag', by converting the APPEND_ADD_WITH_NOP machine code generation to APPEND_ADD where a relaxed instruction is recorded. Relaxation then, depending on actual code produced, chooses between either using a compact branch or jump encoding and emitting the NOP outstanding if no compact encoding is possible. For code simplicity's sake the relaxation pass is retained even if the principle of preferring a compact encoding to a 16-bit branch/NOP pair means, in the absence of out-of-range branch relaxation, that a single compact branch machine code instruction will eventually be produced from a given assembly source instruction. gas/ * config/tc-mips.c (RELAX_MICROMIPS_ENCODE): Add `nods' flag. (RELAX_MICROMIPS_RELAX32, RELAX_MICROMIPS_TOOFAR16) (RELAX_MICROMIPS_MARK_TOOFAR16, RELAX_MICROMIPS_CLEAR_TOOFAR16) (RELAX_MICROMIPS_TOOFAR32, RELAX_MICROMIPS_MARK_TOOFAR32) (RELAX_MICROMIPS_CLEAR_TOOFAR32): Shift bits. (get_append_method): Also return APPEND_ADD_COMPACT for microMIPS instructions. (find_altered_mips16_opcode): Exclude macros from matching. Factor code out... (find_altered_opcode): ... to this new function. (find_altered_micromips_opcode): New function. (frag_branch_delay_slot_size): Likewise. (append_insn): Handle microMIPS branch/jump compaction. (macro_start): Likewise. (relaxed_micromips_32bit_branch_length): Likewise. (md_convert_frag): Likewise. * testsuite/gas/mips/micromips.s: Add conditional explicit NOPs for delay slot filling. * testsuite/gas/mips/micromips-b16.s: Add explicit NOPs for delay slot filling. * testsuite/gas/mips/micromips-size-1.s: Likewise. * testsuite/gas/mips/micromips.l: Adjust line numbers. * testsuite/gas/mips/micromips-warn.l: Likewise. * testsuite/gas/mips/micromips-size-1.l: Likewise. * testsuite/gas/mips/micromips.d: Adjust padding. * testsuite/gas/mips/micromips-trap.d: Likewise. * testsuite/gas/mips/micromips-insn32.d: Likewise. * testsuite/gas/mips/micromips-noinsn32.d: Likewise. * testsuite/gas/mips/micromips@beq.d: Update patterns for branch/jump compaction. * testsuite/gas/mips/micromips@bge.d: Likewise. * testsuite/gas/mips/micromips@bgeu.d: Likewise. * testsuite/gas/mips/micromips@blt.d: Likewise. * testsuite/gas/mips/micromips@bltu.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-4.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-4-64.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5pic.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5-64.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5pic-64.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local-n32.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local-n64.d: Likewise. * testsuite/gas/mips/micromips@loc-swap.d: Likewise. * testsuite/gas/mips/micromips@loc-swap-dis.d: Likewise. * testsuite/gas/mips/micromips@relax.d: Likewise. * testsuite/gas/mips/micromips@relax-at.d: Likewise. * testsuite/gas/mips/micromips@relax-swap3.d: Likewise. * testsuite/gas/mips/branch-extern-2.d: Likewise. * testsuite/gas/mips/branch-extern-4.d: Likewise. * testsuite/gas/mips/branch-section-2.d: Likewise. * testsuite/gas/mips/branch-section-4.d: Likewise. * testsuite/gas/mips/branch-weak-2.d: Likewise. * testsuite/gas/mips/branch-weak-5.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-n32.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-n64.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend-n32.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend-n64.d: Likewise. * testsuite/gas/mips/micromips-compact.d: New test. * testsuite/gas/mips/mips.exp: Run the new test. ld/ * testsuite/ld-mips-elf/micromips-branch-absolute.d: Update patterns for branch compaction. * testsuite/ld-mips-elf/micromips-branch-absolute-addend.d: Likewise. opcodes/ * micromips-opc.c (micromips_opcodes): Reorder "bc" next to "b", "beqzc" next to "beq", "bnezc" next to "bne" and "jrc" next to "j".
2016-07-28 00:27:55 +08:00
[0-9a-f]+ <[^>]*> 40e0 0118 bc 0+001234 <foo\+0x234>
[0-9a-f]+ <[^>]*> 4060 0116 bal 0+001234 <foo\+0x234>
BFD: Let targets handle relocations against absolute symbols Fix a generic BFD issue with relocations against absolute symbols, which are installed without using any individual relocation handler provided by the backend. This causes any absolute section's addend to be lost on REL targets such as o32 MIPS, and also relocation-specific calculation adjustments are not made. As an example assembling this program: $ cat test.s .text foo: b bar b baz .set bar, 0x1234 $ as -EB -32 -o test-o32.o test.s $ as -EB -n32 -o test-n32.o test.s produces this binary code: $ objdump -dr test-o32.o test-n32.o test-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop test-n32.o: file format elf32-ntradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS*+0x1230 4: 00000000 nop 8: 10000000 b c <foo+0xc> 8: R_MIPS_PC16 baz-0x4 c: 00000000 nop $ where it is clearly visible in `test-o32.o', which uses REL relocations, that the absolute section's addend equivalent to the value of `bar' -- a reference to which cannot be fully resolved at the assembly time, because the reference is PC-relative -- has been lost, as has been the relocation-specific adjustment of -4, required to take into account the PC+4-relative calculation made by hardware with branches and seen in the external symbol reference to `baz' as the `ffff' addend encoded in the instruction word. In `test-n32.o', which uses RELA relocations, the absolute section's addend has been correctly retained. Give precedence then in `bfd_perform_relocation' and `bfd_install_relocation' to any individual relocation handler the backend selected may have provided, while still resorting to the generic calculation otherwise. This retains the semantics which we've had since forever or before the beginning of our repository history, and is at the very least compatible with `bfd_elf_generic_reloc' being used as the handler. Retain the `bfd_is_und_section' check unchanged at the beginning of `bfd_perform_relocation' since this does not affect the semantics of the function. The check returns the same `bfd_reloc_undefined' code the check for a null `howto' does, so swapping the two does not matter. Also the check is is mutually exclusive with the `bfd_is_abs_section' check, since a section cannot be absolute and undefined both at once, so swapping the two does not matter either. With this change applied the program quoted above now has the in-place addend correctly calculated and installed in the field being relocated: $ objdump -dr fixed-o32.o fixed-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 1000048c b 1234 <bar> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop $ Add a set of MIPS tests to cover the relevant cases, including absolute symbols with addends, and verifying that PC-relative relocations against symbols concerned resolve to the same value in the final link regardless of whether the REL or the RELA relocation form is used. Exclude linker tests though which would overflow the in-place addend on REL targets and use them as dump patterns for RELA targets only. bfd/ * reloc.c (bfd_perform_relocation): Try the `howto' handler first with relocations against absolute symbols. (bfd_install_relocation): Likewise. gas/ * testsuite/gas/mips/mips16-branch-absolute.d: Update patterns. * testsuite/gas/mips/branch-absolute.d: New test. * testsuite/gas/mips/branch-absolute-n32.d: New test. * testsuite/gas/mips/branch-absolute-n64.d: New test. * testsuite/gas/mips/branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/branch-absolute.s: New test source. * testsuite/gas/mips/branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips16-branch-absolute-addend.s: New test source. * testsuite/gas/mips/micromips-branch-absolute.s: New test source. * testsuite/gas/mips/micromips-branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips.exp: Run the new tests. ld/ * testsuite/ld-mips-elf/branch-absolute.d: New test. * testsuite/ld-mips-elf/branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/mips-elf.exp: Run the new tests, except from `branch-absolute-addend' and `micromips-branch-absolute-addend', referred indirectly only.
2016-07-12 08:30:01 +08:00
[0-9a-f]+ <[^>]*> 0000 0000 nop
MIPS/GAS: Implement microMIPS branch/jump compaction Convert microMIPS branches and jumps whose delay slot would be filled by a generated NOP instruction to the corresponding compact form where one exists, in a manner similar to MIPS16 JR->JRC and JALR->JALRC swap. Do so even where the transformation switches from a 16-bit to a 32-bit branch encoding for no benefit in code size reduction, as this is still advantageous. This is because a branch/NOP pair takes 2 pipeline slots or a 2-cycle completion latency except in superscalar implementations. Whereas a compact branch may or may not stall on its target fetch, so it will at most have a 2-cycle completion latency and may have only 1 even in scalar implementations, and in superscalar implementations it is expected to have no worse latency as a branch/NOP pair has. Also it won't stall and therefore take the extra latency cycle in the not-taken case. Technically this is the same as MIPS16 compaction: for the qualifying instruction encodings the APPEND_ADD_COMPACT machine code generation method is selected where APPEND_ADD_WITH_NOP otherwise would and tells the code generator in `append_insn' to convert the regular form of an instruction to its corresponding compact form. For this the opcode is tweaked as necessary and the microMIPS opcode table is scanned for the matching updated instruction. A non-$0 `rt' operand to BEQ and BNE instructions is moved to the `rs' operand field of BEQZC and BNEZC encodings as required. Unlike with MIPS16 compaction however we need to handle out-of-distance branch relaxation as well. We do this by deferring the generation of any delay-slot NOP required to relaxation made in `md_convert_frag', by converting the APPEND_ADD_WITH_NOP machine code generation to APPEND_ADD where a relaxed instruction is recorded. Relaxation then, depending on actual code produced, chooses between either using a compact branch or jump encoding and emitting the NOP outstanding if no compact encoding is possible. For code simplicity's sake the relaxation pass is retained even if the principle of preferring a compact encoding to a 16-bit branch/NOP pair means, in the absence of out-of-range branch relaxation, that a single compact branch machine code instruction will eventually be produced from a given assembly source instruction. gas/ * config/tc-mips.c (RELAX_MICROMIPS_ENCODE): Add `nods' flag. (RELAX_MICROMIPS_RELAX32, RELAX_MICROMIPS_TOOFAR16) (RELAX_MICROMIPS_MARK_TOOFAR16, RELAX_MICROMIPS_CLEAR_TOOFAR16) (RELAX_MICROMIPS_TOOFAR32, RELAX_MICROMIPS_MARK_TOOFAR32) (RELAX_MICROMIPS_CLEAR_TOOFAR32): Shift bits. (get_append_method): Also return APPEND_ADD_COMPACT for microMIPS instructions. (find_altered_mips16_opcode): Exclude macros from matching. Factor code out... (find_altered_opcode): ... to this new function. (find_altered_micromips_opcode): New function. (frag_branch_delay_slot_size): Likewise. (append_insn): Handle microMIPS branch/jump compaction. (macro_start): Likewise. (relaxed_micromips_32bit_branch_length): Likewise. (md_convert_frag): Likewise. * testsuite/gas/mips/micromips.s: Add conditional explicit NOPs for delay slot filling. * testsuite/gas/mips/micromips-b16.s: Add explicit NOPs for delay slot filling. * testsuite/gas/mips/micromips-size-1.s: Likewise. * testsuite/gas/mips/micromips.l: Adjust line numbers. * testsuite/gas/mips/micromips-warn.l: Likewise. * testsuite/gas/mips/micromips-size-1.l: Likewise. * testsuite/gas/mips/micromips.d: Adjust padding. * testsuite/gas/mips/micromips-trap.d: Likewise. * testsuite/gas/mips/micromips-insn32.d: Likewise. * testsuite/gas/mips/micromips-noinsn32.d: Likewise. * testsuite/gas/mips/micromips@beq.d: Update patterns for branch/jump compaction. * testsuite/gas/mips/micromips@bge.d: Likewise. * testsuite/gas/mips/micromips@bgeu.d: Likewise. * testsuite/gas/mips/micromips@blt.d: Likewise. * testsuite/gas/mips/micromips@bltu.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-4.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-4-64.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5pic.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5-64.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5pic-64.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local-n32.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local-n64.d: Likewise. * testsuite/gas/mips/micromips@loc-swap.d: Likewise. * testsuite/gas/mips/micromips@loc-swap-dis.d: Likewise. * testsuite/gas/mips/micromips@relax.d: Likewise. * testsuite/gas/mips/micromips@relax-at.d: Likewise. * testsuite/gas/mips/micromips@relax-swap3.d: Likewise. * testsuite/gas/mips/branch-extern-2.d: Likewise. * testsuite/gas/mips/branch-extern-4.d: Likewise. * testsuite/gas/mips/branch-section-2.d: Likewise. * testsuite/gas/mips/branch-section-4.d: Likewise. * testsuite/gas/mips/branch-weak-2.d: Likewise. * testsuite/gas/mips/branch-weak-5.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-n32.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-n64.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend-n32.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend-n64.d: Likewise. * testsuite/gas/mips/micromips-compact.d: New test. * testsuite/gas/mips/mips.exp: Run the new test. ld/ * testsuite/ld-mips-elf/micromips-branch-absolute.d: Update patterns for branch compaction. * testsuite/ld-mips-elf/micromips-branch-absolute-addend.d: Likewise. opcodes/ * micromips-opc.c (micromips_opcodes): Reorder "bc" next to "b", "beqzc" next to "beq", "bnezc" next to "bne" and "jrc" next to "j".
2016-07-28 00:27:55 +08:00
[0-9a-f]+ <[^>]*> 4020 0112 bltzal zero,0+001234 <foo\+0x234>
BFD: Let targets handle relocations against absolute symbols Fix a generic BFD issue with relocations against absolute symbols, which are installed without using any individual relocation handler provided by the backend. This causes any absolute section's addend to be lost on REL targets such as o32 MIPS, and also relocation-specific calculation adjustments are not made. As an example assembling this program: $ cat test.s .text foo: b bar b baz .set bar, 0x1234 $ as -EB -32 -o test-o32.o test.s $ as -EB -n32 -o test-n32.o test.s produces this binary code: $ objdump -dr test-o32.o test-n32.o test-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop test-n32.o: file format elf32-ntradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS*+0x1230 4: 00000000 nop 8: 10000000 b c <foo+0xc> 8: R_MIPS_PC16 baz-0x4 c: 00000000 nop $ where it is clearly visible in `test-o32.o', which uses REL relocations, that the absolute section's addend equivalent to the value of `bar' -- a reference to which cannot be fully resolved at the assembly time, because the reference is PC-relative -- has been lost, as has been the relocation-specific adjustment of -4, required to take into account the PC+4-relative calculation made by hardware with branches and seen in the external symbol reference to `baz' as the `ffff' addend encoded in the instruction word. In `test-n32.o', which uses RELA relocations, the absolute section's addend has been correctly retained. Give precedence then in `bfd_perform_relocation' and `bfd_install_relocation' to any individual relocation handler the backend selected may have provided, while still resorting to the generic calculation otherwise. This retains the semantics which we've had since forever or before the beginning of our repository history, and is at the very least compatible with `bfd_elf_generic_reloc' being used as the handler. Retain the `bfd_is_und_section' check unchanged at the beginning of `bfd_perform_relocation' since this does not affect the semantics of the function. The check returns the same `bfd_reloc_undefined' code the check for a null `howto' does, so swapping the two does not matter. Also the check is is mutually exclusive with the `bfd_is_abs_section' check, since a section cannot be absolute and undefined both at once, so swapping the two does not matter either. With this change applied the program quoted above now has the in-place addend correctly calculated and installed in the field being relocated: $ objdump -dr fixed-o32.o fixed-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 1000048c b 1234 <bar> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop $ Add a set of MIPS tests to cover the relevant cases, including absolute symbols with addends, and verifying that PC-relative relocations against symbols concerned resolve to the same value in the final link regardless of whether the REL or the RELA relocation form is used. Exclude linker tests though which would overflow the in-place addend on REL targets and use them as dump patterns for RELA targets only. bfd/ * reloc.c (bfd_perform_relocation): Try the `howto' handler first with relocations against absolute symbols. (bfd_install_relocation): Likewise. gas/ * testsuite/gas/mips/mips16-branch-absolute.d: Update patterns. * testsuite/gas/mips/branch-absolute.d: New test. * testsuite/gas/mips/branch-absolute-n32.d: New test. * testsuite/gas/mips/branch-absolute-n64.d: New test. * testsuite/gas/mips/branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/branch-absolute.s: New test source. * testsuite/gas/mips/branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips16-branch-absolute-addend.s: New test source. * testsuite/gas/mips/micromips-branch-absolute.s: New test source. * testsuite/gas/mips/micromips-branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips.exp: Run the new tests. ld/ * testsuite/ld-mips-elf/branch-absolute.d: New test. * testsuite/ld-mips-elf/branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/mips-elf.exp: Run the new tests, except from `branch-absolute-addend' and `micromips-branch-absolute-addend', referred indirectly only.
2016-07-12 08:30:01 +08:00
[0-9a-f]+ <[^>]*> 0000 0000 nop
MIPS/GAS: Implement microMIPS branch/jump compaction Convert microMIPS branches and jumps whose delay slot would be filled by a generated NOP instruction to the corresponding compact form where one exists, in a manner similar to MIPS16 JR->JRC and JALR->JALRC swap. Do so even where the transformation switches from a 16-bit to a 32-bit branch encoding for no benefit in code size reduction, as this is still advantageous. This is because a branch/NOP pair takes 2 pipeline slots or a 2-cycle completion latency except in superscalar implementations. Whereas a compact branch may or may not stall on its target fetch, so it will at most have a 2-cycle completion latency and may have only 1 even in scalar implementations, and in superscalar implementations it is expected to have no worse latency as a branch/NOP pair has. Also it won't stall and therefore take the extra latency cycle in the not-taken case. Technically this is the same as MIPS16 compaction: for the qualifying instruction encodings the APPEND_ADD_COMPACT machine code generation method is selected where APPEND_ADD_WITH_NOP otherwise would and tells the code generator in `append_insn' to convert the regular form of an instruction to its corresponding compact form. For this the opcode is tweaked as necessary and the microMIPS opcode table is scanned for the matching updated instruction. A non-$0 `rt' operand to BEQ and BNE instructions is moved to the `rs' operand field of BEQZC and BNEZC encodings as required. Unlike with MIPS16 compaction however we need to handle out-of-distance branch relaxation as well. We do this by deferring the generation of any delay-slot NOP required to relaxation made in `md_convert_frag', by converting the APPEND_ADD_WITH_NOP machine code generation to APPEND_ADD where a relaxed instruction is recorded. Relaxation then, depending on actual code produced, chooses between either using a compact branch or jump encoding and emitting the NOP outstanding if no compact encoding is possible. For code simplicity's sake the relaxation pass is retained even if the principle of preferring a compact encoding to a 16-bit branch/NOP pair means, in the absence of out-of-range branch relaxation, that a single compact branch machine code instruction will eventually be produced from a given assembly source instruction. gas/ * config/tc-mips.c (RELAX_MICROMIPS_ENCODE): Add `nods' flag. (RELAX_MICROMIPS_RELAX32, RELAX_MICROMIPS_TOOFAR16) (RELAX_MICROMIPS_MARK_TOOFAR16, RELAX_MICROMIPS_CLEAR_TOOFAR16) (RELAX_MICROMIPS_TOOFAR32, RELAX_MICROMIPS_MARK_TOOFAR32) (RELAX_MICROMIPS_CLEAR_TOOFAR32): Shift bits. (get_append_method): Also return APPEND_ADD_COMPACT for microMIPS instructions. (find_altered_mips16_opcode): Exclude macros from matching. Factor code out... (find_altered_opcode): ... to this new function. (find_altered_micromips_opcode): New function. (frag_branch_delay_slot_size): Likewise. (append_insn): Handle microMIPS branch/jump compaction. (macro_start): Likewise. (relaxed_micromips_32bit_branch_length): Likewise. (md_convert_frag): Likewise. * testsuite/gas/mips/micromips.s: Add conditional explicit NOPs for delay slot filling. * testsuite/gas/mips/micromips-b16.s: Add explicit NOPs for delay slot filling. * testsuite/gas/mips/micromips-size-1.s: Likewise. * testsuite/gas/mips/micromips.l: Adjust line numbers. * testsuite/gas/mips/micromips-warn.l: Likewise. * testsuite/gas/mips/micromips-size-1.l: Likewise. * testsuite/gas/mips/micromips.d: Adjust padding. * testsuite/gas/mips/micromips-trap.d: Likewise. * testsuite/gas/mips/micromips-insn32.d: Likewise. * testsuite/gas/mips/micromips-noinsn32.d: Likewise. * testsuite/gas/mips/micromips@beq.d: Update patterns for branch/jump compaction. * testsuite/gas/mips/micromips@bge.d: Likewise. * testsuite/gas/mips/micromips@bgeu.d: Likewise. * testsuite/gas/mips/micromips@blt.d: Likewise. * testsuite/gas/mips/micromips@bltu.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-4.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-4-64.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5pic.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5-64.d: Likewise. * testsuite/gas/mips/micromips@branch-misc-5pic-64.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local-n32.d: Likewise. * testsuite/gas/mips/micromips@jal-svr4pic-local-n64.d: Likewise. * testsuite/gas/mips/micromips@loc-swap.d: Likewise. * testsuite/gas/mips/micromips@loc-swap-dis.d: Likewise. * testsuite/gas/mips/micromips@relax.d: Likewise. * testsuite/gas/mips/micromips@relax-at.d: Likewise. * testsuite/gas/mips/micromips@relax-swap3.d: Likewise. * testsuite/gas/mips/branch-extern-2.d: Likewise. * testsuite/gas/mips/branch-extern-4.d: Likewise. * testsuite/gas/mips/branch-section-2.d: Likewise. * testsuite/gas/mips/branch-section-4.d: Likewise. * testsuite/gas/mips/branch-weak-2.d: Likewise. * testsuite/gas/mips/branch-weak-5.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-n32.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-n64.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend-n32.d: Likewise. * testsuite/gas/mips/micromips-branch-absolute-addend-n64.d: Likewise. * testsuite/gas/mips/micromips-compact.d: New test. * testsuite/gas/mips/mips.exp: Run the new test. ld/ * testsuite/ld-mips-elf/micromips-branch-absolute.d: Update patterns for branch compaction. * testsuite/ld-mips-elf/micromips-branch-absolute-addend.d: Likewise. opcodes/ * micromips-opc.c (micromips_opcodes): Reorder "bc" next to "b", "beqzc" next to "beq", "bnezc" next to "bne" and "jrc" next to "j".
2016-07-28 00:27:55 +08:00
[0-9a-f]+ <[^>]*> 40e2 010e beqzc v0,0+001234 <foo\+0x234>
[0-9a-f]+ <[^>]*> 40a2 010c bnezc v0,0+001234 <foo\+0x234>
BFD: Let targets handle relocations against absolute symbols Fix a generic BFD issue with relocations against absolute symbols, which are installed without using any individual relocation handler provided by the backend. This causes any absolute section's addend to be lost on REL targets such as o32 MIPS, and also relocation-specific calculation adjustments are not made. As an example assembling this program: $ cat test.s .text foo: b bar b baz .set bar, 0x1234 $ as -EB -32 -o test-o32.o test.s $ as -EB -n32 -o test-n32.o test.s produces this binary code: $ objdump -dr test-o32.o test-n32.o test-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop test-n32.o: file format elf32-ntradbigmips Disassembly of section .text: 00000000 <foo>: 0: 10000000 b 4 <foo+0x4> 0: R_MIPS_PC16 *ABS*+0x1230 4: 00000000 nop 8: 10000000 b c <foo+0xc> 8: R_MIPS_PC16 baz-0x4 c: 00000000 nop $ where it is clearly visible in `test-o32.o', which uses REL relocations, that the absolute section's addend equivalent to the value of `bar' -- a reference to which cannot be fully resolved at the assembly time, because the reference is PC-relative -- has been lost, as has been the relocation-specific adjustment of -4, required to take into account the PC+4-relative calculation made by hardware with branches and seen in the external symbol reference to `baz' as the `ffff' addend encoded in the instruction word. In `test-n32.o', which uses RELA relocations, the absolute section's addend has been correctly retained. Give precedence then in `bfd_perform_relocation' and `bfd_install_relocation' to any individual relocation handler the backend selected may have provided, while still resorting to the generic calculation otherwise. This retains the semantics which we've had since forever or before the beginning of our repository history, and is at the very least compatible with `bfd_elf_generic_reloc' being used as the handler. Retain the `bfd_is_und_section' check unchanged at the beginning of `bfd_perform_relocation' since this does not affect the semantics of the function. The check returns the same `bfd_reloc_undefined' code the check for a null `howto' does, so swapping the two does not matter. Also the check is is mutually exclusive with the `bfd_is_abs_section' check, since a section cannot be absolute and undefined both at once, so swapping the two does not matter either. With this change applied the program quoted above now has the in-place addend correctly calculated and installed in the field being relocated: $ objdump -dr fixed-o32.o fixed-o32.o: file format elf32-tradbigmips Disassembly of section .text: 00000000 <foo>: 0: 1000048c b 1234 <bar> 0: R_MIPS_PC16 *ABS* 4: 00000000 nop 8: 1000ffff b 8 <foo+0x8> 8: R_MIPS_PC16 baz c: 00000000 nop $ Add a set of MIPS tests to cover the relevant cases, including absolute symbols with addends, and verifying that PC-relative relocations against symbols concerned resolve to the same value in the final link regardless of whether the REL or the RELA relocation form is used. Exclude linker tests though which would overflow the in-place addend on REL targets and use them as dump patterns for RELA targets only. bfd/ * reloc.c (bfd_perform_relocation): Try the `howto' handler first with relocations against absolute symbols. (bfd_install_relocation): Likewise. gas/ * testsuite/gas/mips/mips16-branch-absolute.d: Update patterns. * testsuite/gas/mips/branch-absolute.d: New test. * testsuite/gas/mips/branch-absolute-n32.d: New test. * testsuite/gas/mips/branch-absolute-n64.d: New test. * testsuite/gas/mips/branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-n64.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/mips16-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-n64.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n32.d: New test. * testsuite/gas/mips/micromips-branch-absolute-addend-n64.d: New test. * testsuite/gas/mips/branch-absolute.s: New test source. * testsuite/gas/mips/branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips16-branch-absolute-addend.s: New test source. * testsuite/gas/mips/micromips-branch-absolute.s: New test source. * testsuite/gas/mips/micromips-branch-absolute-addend.s: New test source. * testsuite/gas/mips/mips.exp: Run the new tests. ld/ * testsuite/ld-mips-elf/branch-absolute.d: New test. * testsuite/ld-mips-elf/branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-n64.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n32.d: New test. * testsuite/ld-mips-elf/micromips-branch-absolute-addend-n64.d: New test. * testsuite/ld-mips-elf/mips-elf.exp: Run the new tests, except from `branch-absolute-addend' and `micromips-branch-absolute-addend', referred indirectly only.
2016-07-12 08:30:01 +08:00
[0-9a-f]+ <[^>]*> 0c00 nop
\.\.\.