summaryrefslogtreecommitdiff
path: root/tools/objtool/arch
AgeCommit message (Collapse)AuthorFilesLines
2023-06-28Merge tag 'objtool-core-2023-06-27' of ↵Linus Torvalds4-13/+21
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull objtool updates from Ingo Molar: "Build footprint & performance improvements: - Reduce memory usage with CONFIG_DEBUG_INFO=y In the worst case of an allyesconfig+CONFIG_DEBUG_INFO=y kernel, DWARF creates almost 200 million relocations, ballooning objtool's peak heap usage to 53GB. These patches reduce that to 25GB. On a distro-type kernel with kernel IBT enabled, they reduce objtool's peak heap usage from 4.2GB to 2.8GB. These changes also improve the runtime significantly. Debuggability improvements: - Add the unwind_debug command-line option, for more extend unwinding debugging output - Limit unreachable warnings to once per function - Add verbose option for disassembling affected functions - Include backtrace in verbose mode - Detect missing __noreturn annotations - Ignore exc_double_fault() __noreturn warnings - Remove superfluous global_noreturns entries - Move noreturn function list to separate file - Add __kunit_abort() to noreturns Unwinder improvements: - Allow stack operations in UNWIND_HINT_UNDEFINED regions - drm/vmwgfx: Add unwind hints around RBP clobber Cleanups: - Move the x86 entry thunk restore code into thunk functions - x86/unwind/orc: Use swap() instead of open coding it - Remove unnecessary/unused variables Fixes for modern stack canary handling" * tag 'objtool-core-2023-06-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (42 commits) x86/orc: Make the is_callthunk() definition depend on CONFIG_BPF_JIT=y objtool: Skip reading DWARF section data objtool: Free insns when done objtool: Get rid of reloc->rel[a] objtool: Shrink elf hash nodes objtool: Shrink reloc->sym_reloc_entry objtool: Get rid of reloc->jump_table_start objtool: Get rid of reloc->addend objtool: Get rid of reloc->type objtool: Get rid of reloc->offset objtool: Get rid of reloc->idx objtool: Get rid of reloc->list objtool: Allocate relocs in advance for new rela sections objtool: Add for_each_reloc() objtool: Don't free memory in elf_close() objtool: Keep GElf_Rel[a] structs synced objtool: Add elf_create_section_pair() objtool: Add mark_sec_changed() objtool: Fix reloc_hash size objtool: Consolidate rel/rela handling ...
2023-06-07objtool: Get rid of reloc->addendJosh Poimboeuf2-3/+3
Get the addend from the embedded GElf_Rel[a] struct. With allyesconfig + CONFIG_DEBUG_INFO: - Before: peak heap memory consumption: 42.10G - After: peak heap memory consumption: 40.37G Link: https://lore.kernel.org/r/ad2354f95d9ddd86094e3f7687acfa0750657784.1685464332.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-06-07objtool: Get rid of reloc->typeJosh Poimboeuf2-3/+3
Get the type from the embedded GElf_Rel[a] struct. Link: https://lore.kernel.org/r/d1c1f8da31e4f052a2478aea585fcf355cacc53a.1685464332.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-06-07objtool: Add elf_create_section_pair()Josh Poimboeuf2-7/+15
When creating an annotation section, allocate the reloc section data at the beginning. This simplifies the data model a bit and also saves memory due to the removal of malloc() in elf_rebuild_reloc_section(). With allyesconfig + CONFIG_DEBUG_INFO: - Before: peak heap memory consumption: 53.49G - After: peak heap memory consumption: 49.02G Link: https://lore.kernel.org/r/048e908f3ede9b66c15e44672b6dda992b1dae3e.1685464332.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-05-10x86/alternative: Support relocations in alternativesPeter Zijlstra1-7/+1
A little while ago someone (Kirill) ran into the whole 'alternatives don't do relocations nonsense' again and I got annoyed enough to actually look at the code. Since the whole alternative machinery already fully decodes the instructions it is simple enough to adjust immediates and displacement when needed. Specifically, the immediates for IP modifying instructions (JMP, CALL, Jcc) and the displacement for RIP-relative instructions. [ bp: Massage comment some more and get rid of third loop in apply_relocation(). ] Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de> Link: https://lore.kernel.org/r/20230208171431.313857925@infradead.org
2023-02-23objtool: Make instruction::stack_ops a single-linked listPeter Zijlstra1-2/+2
struct instruction { struct list_head list; /* 0 16 */ struct hlist_node hash; /* 16 16 */ struct list_head call_node; /* 32 16 */ struct section * sec; /* 48 8 */ long unsigned int offset; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ unsigned int len; /* 64 4 */ enum insn_type type; /* 68 4 */ long unsigned int immediate; /* 72 8 */ u16 dead_end:1; /* 80: 0 2 */ u16 ignore:1; /* 80: 1 2 */ u16 ignore_alts:1; /* 80: 2 2 */ u16 hint:1; /* 80: 3 2 */ u16 save:1; /* 80: 4 2 */ u16 restore:1; /* 80: 5 2 */ u16 retpoline_safe:1; /* 80: 6 2 */ u16 noendbr:1; /* 80: 7 2 */ u16 entry:1; /* 80: 8 2 */ /* XXX 7 bits hole, try to pack */ s8 instr; /* 82 1 */ u8 visited; /* 83 1 */ /* XXX 4 bytes hole, try to pack */ struct alt_group * alt_group; /* 88 8 */ struct symbol * call_dest; /* 96 8 */ struct instruction * jump_dest; /* 104 8 */ struct instruction * first_jump_src; /* 112 8 */ struct reloc * jump_table; /* 120 8 */ /* --- cacheline 2 boundary (128 bytes) --- */ struct reloc * reloc; /* 128 8 */ struct list_head alts; /* 136 16 */ struct symbol * sym; /* 152 8 */ - struct list_head stack_ops; /* 160 16 */ - struct cfi_state * cfi; /* 176 8 */ + struct stack_op * stack_ops; /* 160 8 */ + struct cfi_state * cfi; /* 168 8 */ - /* size: 184, cachelines: 3, members: 29 */ - /* sum members: 178, holes: 1, sum holes: 4 */ + /* size: 176, cachelines: 3, members: 29 */ + /* sum members: 170, holes: 1, sum holes: 4 */ /* sum bitfield members: 9 bits, bit holes: 1, sum bit holes: 7 bits */ - /* last cacheline: 56 bytes */ + /* last cacheline: 48 bytes */ }; pre: 5:58.22 real, 226.69 user, 131.22 sys, 26221520 mem post: 5:58.50 real, 229.64 user, 128.65 sys, 26221520 mem Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Tested-by: Nathan Chancellor <nathan@kernel.org> # build only Tested-by: Thomas Weißschuh <linux@weissschuh.net> # compile and run Link: https://lore.kernel.org/r/20230208172245.362196959@infradead.org
2023-02-23objtool: Change arch_decode_instruction() signaturePeter Zijlstra2-65/+62
In preparation to changing struct instruction around a bit, avoid passing it's members by pointer and instead pass the whole thing. A cleanup in it's own right too. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Tested-by: Nathan Chancellor <nathan@kernel.org> # build only Tested-by: Thomas Weißschuh <linux@weissschuh.net> # compile and run Link: https://lore.kernel.org/r/20230208172245.291087549@infradead.org
2023-01-05x86/alternatives: Add alt_instr.flagsBorislav Petkov (AMD)1-3/+3
Add a struct alt_instr.flags field which will contain different flags controlling alternatives patching behavior. The initial idea was to be able to specify it as a separate macro parameter but that would mean touching all possible invocations of the alternatives macros and thus a lot of churn. What is more, as PeterZ suggested, being able to say ALT_NOT(feature) is very readable and explains exactly what is meant. So make the feature field a u32 where the patching flags are the upper u16 part of the dword quantity while the lower u16 word is the feature. The highest feature number currently is 0x26a (i.e., word 19) so there is plenty of space. If that becomes insufficient, the field can be extended to u64 which will then make struct alt_instr of the nice size of 16 bytes (14 bytes currently). There should be no functional changes resulting from this. Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de> Reviewed-by: Ingo Molnar <mingo@kernel.org> Link: https://lore.kernel.org/r/Y6RCoJEtxxZWwotd@zn.tnic
2022-12-19Merge tag 'powerpc-6.2-1' of ↵Linus Torvalds9-9/+180
git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux Pull powerpc updates from Michael Ellerman: - Add powerpc qspinlock implementation optimised for large system scalability and paravirt. See the merge message for more details - Enable objtool to be built on powerpc to generate mcount locations - Use a temporary mm for code patching with the Radix MMU, so the writable mapping is restricted to the patching CPU - Add an option to build the 64-bit big-endian kernel with the ELFv2 ABI - Sanitise user registers on interrupt entry on 64-bit Book3S - Many other small features and fixes Thanks to Aboorva Devarajan, Angel Iglesias, Benjamin Gray, Bjorn Helgaas, Bo Liu, Chen Lifu, Christoph Hellwig, Christophe JAILLET, Christophe Leroy, Christopher M. Riedl, Colin Ian King, Deming Wang, Disha Goel, Dmitry Torokhov, Finn Thain, Geert Uytterhoeven, Gustavo A. R. Silva, Haowen Bai, Joel Stanley, Jordan Niethe, Julia Lawall, Kajol Jain, Laurent Dufour, Li zeming, Miaoqian Lin, Michael Jeanson, Nathan Lynch, Naveen N. Rao, Nayna Jain, Nicholas Miehlbradt, Nicholas Piggin, Pali Rohár, Randy Dunlap, Rohan McLure, Russell Currey, Sathvika Vasireddy, Shaomin Deng, Stephen Kitt, Stephen Rothwell, Thomas Weißschuh, Tiezhu Yang, Uwe Kleine-König, Xie Shaowen, Xiu Jianfeng, XueBing Chen, Yang Yingliang, Zhang Jiaming, ruanjinjie, Jessica Yu, and Wolfram Sang. * tag 'powerpc-6.2-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (181 commits) powerpc/code-patching: Fix oops with DEBUG_VM enabled powerpc/qspinlock: Fix 32-bit build powerpc/prom: Fix 32-bit build powerpc/rtas: mandate RTAS syscall filtering powerpc/rtas: define pr_fmt and convert printk call sites powerpc/rtas: clean up includes powerpc/rtas: clean up rtas_error_log_max initialization powerpc/pseries/eeh: use correct API for error log size powerpc/rtas: avoid scheduling in rtas_os_term() powerpc/rtas: avoid device tree lookups in rtas_os_term() powerpc/rtasd: use correct OF API for event scan rate powerpc/rtas: document rtas_call() powerpc/pseries: unregister VPA when hot unplugging a CPU powerpc/pseries: reset the RCU watchdogs after a LPM powerpc: Take in account addition CPU node when building kexec FDT powerpc: export the CPU node count powerpc/cpuidle: Set CPUIDLE_FLAG_POLLING for snooze state powerpc/dts/fsl: Fix pca954x i2c-mux node names cxl: Remove unnecessary cxl_pci_window_alignment() selftests/powerpc: Fix resource leaks ...
2022-11-23objtool/powerpc: Implement arch_pc_relative_reloc()Michael Ellerman1-0/+9
Provide an implementation for arch_pc_relative_reloc(). It is needed to pass the build once 61c6065ef7ec ("objtool: Allow !PC relative relocations") is merged. Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2022-11-18objtool/powerpc: Add --mcount specific implementationSathvika Vasireddy2-0/+18
This patch enables objtool --mcount on powerpc, and adds implementation specific to powerpc. Tested-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Sathvika Vasireddy <sv@linux.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20221114175754.1131267-17-sv@linux.ibm.com
2022-11-18objtool/powerpc: Enable objtool to be built on ppcSathvika Vasireddy6-0/+146
This patch adds [stub] implementations for required functions, inorder to enable objtool build on powerpc. [Christophe Leroy: powerpc: Add missing asm/asm.h for objtool, Use local variables for type and imm in arch_decode_instruction(), Adapt len for prefixed instructions.] Tested-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Sathvika Vasireddy <sv@linux.ibm.com> Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20221114175754.1131267-16-sv@linux.ibm.com
2022-11-18objtool: Add arch specific function arch_ftrace_match()Sathvika Vasireddy1-0/+5
Add architecture specific function to look for relocation records pointing to architecture specific symbols. Suggested-by: Christophe Leroy <christophe.leroy@csgroup.eu> Tested-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Sathvika Vasireddy <sv@linux.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20221114175754.1131267-15-sv@linux.ibm.com
2022-11-18objtool: Use macros to define arch specific reloc typesSathvika Vasireddy1-0/+2
Make relocation types architecture specific. Tested-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Sathvika Vasireddy <sv@linux.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20221114175754.1131267-14-sv@linux.ibm.com
2022-11-18objtool: Use target file endianness instead of a compiled constantChristophe Leroy1-9/+0
Some architectures like powerpc support both endianness, it's therefore not possible to fix the endianness via arch/endianness.h because there is no easy way to get the target endianness at build time. Use the endianness recorded in the file objtool is working on. Tested-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20221114175754.1131267-10-sv@linux.ibm.com
2022-10-17objtool: Allow !PC relative relocationsPeter Zijlstra1-0/+24
Objtool doesn't currently much like per-cpu usage in alternatives: arch/x86/entry/entry_64.o: warning: objtool: .altinstr_replacement+0xf: unsupported relocation in alternatives section f: 65 c7 04 25 00 00 00 00 00 00 00 80 movl $0x80000000,%gs:0x0 13: R_X86_64_32S __x86_call_depth Since the R_X86_64_32S relocation is location invariant (it's computation doesn't include P - the address of the location itself), it can be trivially allowed. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/20220915111145.806607235@infradead.org
2022-09-15objtool,x86: Teach decode about LOOP* instructionsPeter Zijlstra1-0/+6
When 'discussing' control flow Masami mentioned the LOOP* instructions and I realized objtool doesn't decode them properly. As it turns out, these instructions are somewhat inefficient and as such unlikely to be emitted by the compiler (a few vmlinux.o checks can't find a single one) so this isn't critical, but still, best to decode them properly. Reported-by: Masami Hiramatsu (Google) <mhiramat@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/Yxhd4EMKyoFoH9y4@hirez.programming.kicks-ass.net
2022-06-27x86,objtool: Create .return_sitesPeter Zijlstra1-0/+5
Find all the return-thunk sites and record them in a .return_sites section such that the kernel can undo this. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov <bp@suse.de> Reviewed-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Borislav Petkov <bp@suse.de>
2022-04-22objtool: Reorganize cmdline optionsJosh Poimboeuf2-2/+2
Split the existing options into two groups: actions, which actually do something; and options, which modify the actions in some way. Also there's no need to have short flags for all the non-action options. Reserve short flags for the more important actions. While at it: - change a few of the short flags to be more intuitive - make option descriptions more consistently descriptive - sort options in the source like they are when printed - move options to a global struct Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Link: https://lkml.kernel.org/r/9dcaa752f83aca24b1b21f0b0eeb28a0c181c0b0.1650300597.git.jpoimboe@redhat.com
2022-03-15objtool: Add IBT/ENDBR decodingPeter Zijlstra1-5/+29
Intel IBT requires the target of any indirect CALL or JMP instruction to be the ENDBR instruction; optionally it allows those two instructions to have a NOTRACK prefix in order to avoid this requirement. The kernel will not enable the use of NOTRACK, as such any occurence of it in compiler generated code should be flagged. Teach objtool to Decode ENDBR instructions and WARN about NOTRACK prefixes. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lore.kernel.org/r/20220308154319.645963517@infradead.org
2022-03-15tools/objtool: Check for use of the ENQCMD instruction in the kernelFenghua Yu1-1/+10
The ENQCMD instruction implicitly accesses the PASID_MSR to fill in the pasid field of the descriptor being submitted to an accelerator. But there is no precise (and stable across kernel changes) point at which the PASID_MSR is updated from the value for one task to the next. Kernel code that uses accelerators must always use the ENQCMDS instruction which does not access the PASID_MSR. Check for use of the ENQCMD instruction in the kernel and warn on its usage. Signed-off-by: Fenghua Yu <fenghua.yu@intel.com> Signed-off-by: Borislav Petkov <bp@suse.de> Reviewed-by: Tony Luck <tony.luck@intel.com> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lore.kernel.org/r/20220207230254.3342514-11-fenghua.yu@intel.com Signed-off-by: Peter Zijlstra <peterz@infradead.org>
2021-12-08objtool: Add straight-line-speculation validationPeter Zijlstra1-4/+9
Teach objtool to validate the straight-line-speculation constraints: - speculation trap after indirect calls - speculation trap after RET Notable: when an instruction is annotated RETPOLINE_SAFE, indicating speculation isn't a problem, also don't care about sls for that instruction. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov <bp@suse.de> Link: https://lore.kernel.org/r/20211204134908.023037659@infradead.org
2021-10-29objtool,x86: Replace alternatives with .retpoline_sitesPeter Zijlstra1-120/+0
Instead of writing complete alternatives, simply provide a list of all the retpoline thunk calls. Then the kernel is free to do with them as it pleases. Simpler code all-round. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Borislav Petkov <bp@suse.de> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/r/20211026120309.850007165@infradead.org
2021-10-07Merge branch 'objtool/urgent'Peter Zijlstra1-1/+1
Fixup conflicts. # Conflicts: # tools/objtool/check.c
2021-10-05objtool: Make .altinstructions section entry size consistentJoe Lawrence1-1/+1
Commit e31694e0a7a7 ("objtool: Don't make .altinstructions writable") aligned objtool-created and kernel-created .altinstructions section flags, but there remains a minor discrepency in their use of a section entry size: objtool sets one while the kernel build does not. While sh_entsize of sizeof(struct alt_instr) seems intuitive, this small deviation can cause failures with external tooling (kpatch-build). Fix this by creating new .altinstructions sections with sh_entsize of 0 and then later updating sec->sh_size as alternatives are added to the section. An added benefit is avoiding the data descriptor and buffer created by elf_create_section(), but previously unused by elf_add_alternative(). Fixes: 9bc0bb50727c ("objtool/x86: Rewrite retpoline thunk calls") Signed-off-by: Joe Lawrence <joe.lawrence@redhat.com> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lore.kernel.org/r/20210822225037.54620-2-joe.lawrence@redhat.com Cc: Andy Lavr <andy.lavr@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: x86@kernel.org Cc: linux-kernel@vger.kernel.org
2021-09-17objtool: Support pv_opsindirect calls for noinstrPeter Zijlstra1-1/+33
Normally objtool will now follow indirect calls; there is no need. However, this becomes a problem with noinstr validation; if there's an indirect call from noinstr code, we very much need to know it is to another noinstr function. Luckily there aren't many indirect calls in entry code with the obvious exception of paravirt. As such, noinstr validation didn't work with paravirt kernels. In order to track pv_ops[] call targets, objtool reads the static pv_ops[] tables as well as direct assignments to the pv_ops[] array, provided the compiler makes them a single instruction like: bf87: 48 c7 05 00 00 00 00 00 00 00 00 movq $0x0,0x0(%rip) bf92 <xen_init_spinlocks+0x5f> bf8a: R_X86_64_PC32 pv_ops+0x268 There are, as of yet, no warnings for when this goes wrong :/ Using the functions found with the above means, all pv_ops[] calls are now subject to noinstr validation. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/20210624095149.118815755@infradead.org
2021-09-15objtool: Handle __sanitize_cov*() tail callsPeter Zijlstra1-0/+20
Turns out the compilers also generate tail calls to __sanitize_cov*(), make sure to also patch those out in noinstr code. Fixes: 0f1441b44e82 ("objtool: Fix noinstr vs KCOV") Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Marco Elver <elver@google.com> Link: https://lore.kernel.org/r/20210624095147.818783799@infradead.org
2021-09-15objtool: Introduce CFI hashPeter Zijlstra1-11/+9
Andi reported that objtool on vmlinux.o consumes more memory than his system has, leading to horrific performance. This is in part because we keep a struct instruction for every instruction in the file in-memory. Shrink struct instruction by removing the CFI state (which includes full register state) from it and demand allocating it. Given most instructions don't actually change CFI state, there's lots of repetition there, so add a hash table to find previous CFI instances. Reduces memory consumption (and runtime) for processing an x86_64-allyesconfig: pre: 4:40.84 real, 143.99 user, 44.18 sys, 30624988 mem post: 2:14.61 real, 108.58 user, 25.04 sys, 16396184 mem Suggested-by: Andi Kleen <andi@firstfloor.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/20210624095147.756759107@infradead.org
2021-06-28Merge tags 'objtool-urgent-2021-06-28' and 'objtool-core-2021-06-28' of ↵Linus Torvalds2-1/+2
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull objtool fix and updates from Ingo Molnar: "An ELF format fix for a section flags mismatch bug that breaks kernel tooling such as kpatch-build. The biggest change in this cycle is the new code to handle and rewrite variable sized jump labels - which results in slightly tighter code generation in hot paths, through the use of short(er) NOPs. Also a number of cleanups and fixes, and a change to the generic include/linux/compiler.h to handle a s390 GCC quirk" * tag 'objtool-urgent-2021-06-28' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: objtool: Don't make .altinstructions writable * tag 'objtool-core-2021-06-28' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: objtool: Improve reloc hash size guestimate instrumentation.h: Avoid using inline asm operand modifiers compiler.h: Avoid using inline asm operand modifiers kbuild: Fix objtool dependency for 'OBJECT_FILES_NON_STANDARD_<obj> := n' objtool: Reflow handle_jump_alt() jump_label/x86: Remove unused JUMP_LABEL_NOP_SIZE jump_label, x86: Allow short NOPs objtool: Provide stats for jump_labels objtool: Rewrite jump_label instructions objtool: Decode jump_entry::key addend jump_label, x86: Emit short JMP jump_label: Free jump_entry::key bit1 for build use jump_label, x86: Add variable length patching support jump_label, x86: Introduce jump_entry_size() jump_label, x86: Improve error when we fail expected text jump_label, x86: Factor out the __jump_table generation jump_label, x86: Strip ASM jump_label support x86, objtool: Dont exclude arch/x86/realmode/ objtool: Rewrite hashtable sizing
2021-06-24objtool: Don't make .altinstructions writableJosh Poimboeuf1-1/+1
When objtool creates the .altinstructions section, it sets the SHF_WRITE flag to make the section writable -- unless the section had already been previously created by the kernel. The mismatch between kernel-created and objtool-created section flags can cause failures with external tooling (kpatch-build). And the section doesn't need to be writable anyway. Make the section flags consistent with the kernel's. Fixes: 9bc0bb50727c ("objtool/x86: Rewrite retpoline thunk calls") Reported-by: Joe Lawrence <joe.lawrence@redhat.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Link: https://lore.kernel.org/r/6c284ae89717889ea136f9f0064d914cd8329d31.1624462939.git.jpoimboe@redhat.com
2021-06-11objtool: Only rewrite unconditional retpoline thunk callsPeter Zijlstra1-0/+4
It turns out that the compilers generate conditional branches to the retpoline thunks like: 5d5: 0f 85 00 00 00 00 jne 5db <cpuidle_reflect+0x22> 5d7: R_X86_64_PLT32 __x86_indirect_thunk_r11-0x4 while the rewrite can only handle JMP/CALL to the thunks. The result is the alternative wrecking the code. Make sure to skip writing the alternatives for conditional branches. Fixes: 9bc0bb50727c ("objtool/x86: Rewrite retpoline thunk calls") Reported-by: Lukasz Majczak <lma@semihalf.com> Reported-by: Nathan Chancellor <nathan@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Tested-by: Nathan Chancellor <nathan@kernel.org>
2021-05-12objtool/x86: Fix elf_add_alternative() endiannessVasily Gorbik1-1/+2
Currently x86 kernel cross-compiled on big endian system fails at boot with: kernel BUG at arch/x86/kernel/alternative.c:258! Corresponding bug condition look like the following: BUG_ON(feature >= (NCAPINTS + NBUGINTS) * 32); Fix that by converting alternative feature/cpuid to target endianness. Fixes: 9bc0bb50727c ("objtool/x86: Rewrite retpoline thunk calls") Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: https://lore.kernel.org/r/patch-2.thread-6c9df9.git-6c9df9a8098d.your-ad-here.call-01620841104-ext-2554@work.hours
2021-05-12objtool: Decode jump_entry::key addendPeter Zijlstra1-0/+1
Teach objtool about the the low bits in the struct static_key pointer. That is, the low two bits of @key in: struct jump_entry { s32 code; s32 target; long key; } as found in the __jump_table section. Since @key has a relocation to the variable (to be resolved by the linker), the low two bits will be reflected in the relocation's addend. As such, find the reloc and store the addend, such that we can access these bits. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Link: https://lore.kernel.org/r/20210506194158.028024143@infradead.org
2021-04-28Merge tag 'objtool-core-2021-04-28' of ↵Linus Torvalds2-125/+169
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull objtool updates from Ingo Molnar: - Standardize the crypto asm code so that it looks like compiler- generated code to objtool - so that it can understand it. This enables unwinding from crypto asm code - and also fixes the last known remaining objtool warnings for LTO and more. - x86 decoder fixes: clean up and fix the decoder, and also extend it a bit - Misc fixes and cleanups * tag 'objtool-core-2021-04-28' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (24 commits) x86/crypto: Enable objtool in crypto code x86/crypto/sha512-ssse3: Standardize stack alignment prologue x86/crypto/sha512-avx2: Standardize stack alignment prologue x86/crypto/sha512-avx: Standardize stack alignment prologue x86/crypto/sha256-avx2: Standardize stack alignment prologue x86/crypto/sha1_avx2: Standardize stack alignment prologue x86/crypto/sha_ni: Standardize stack alignment prologue x86/crypto/crc32c-pcl-intel: Standardize jump table x86/crypto/camellia-aesni-avx2: Unconditionally allocate stack buffer x86/crypto/aesni-intel_avx: Standardize stack alignment prologue x86/crypto/aesni-intel_avx: Fix register usage comments x86/crypto/aesni-intel_avx: Remove unused macros objtool: Support asm jump tables objtool: Parse options from OBJTOOL_ARGS objtool: Collate parse_options() users objtool: Add --backup objtool,x86: More ModRM sugar objtool,x86: Rewrite ADD/SUB/AND objtool,x86: Support %riz encodings objtool,x86: Simplify register decode ...
2021-04-02objtool/x86: Rewrite retpoline thunk callsPeter Zijlstra1-0/+117
When the compiler emits: "CALL __x86_indirect_thunk_\reg" for an indirect call, have objtool rewrite it to: ALTERNATIVE "call __x86_indirect_thunk_\reg", "call *%reg", ALT_NOT(X86_FEATURE_RETPOLINE) Additionally, in order to not emit endless identical .altinst_replacement chunks, use a global symbol for them, see __x86_indirect_alt_*. This also avoids objtool from having to do code generation. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov <bp@suse.de> Signed-off-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Link: https://lkml.kernel.org/r/20210326151300.320177914@infradead.org
2021-04-02objtool: Handle per arch retpoline namingPeter Zijlstra1-0/+5
The __x86_indirect_ naming is obviously not generic. Shorten to allow matching some additional magic names later. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov <bp@suse.de> Signed-off-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Link: https://lkml.kernel.org/r/20210326151259.630296706@infradead.org
2021-04-02x86/alternatives: Optimize optimize_nops()Peter Zijlstra1-1/+1
Currently, optimize_nops() scans to see if the alternative starts with NOPs. However, the emit pattern is: 141: \oldinstr 142: .skip (len-(142b-141b)), 0x90 That is, when 'oldinstr' is short, the tail is padded with NOPs. This case never gets optimized. Rewrite optimize_nops() to replace any trailing string of NOPs inside the alternative to larger NOPs. Also run it irrespective of patching, replacing NOPs in both the original and replaced code. A direct consequence is that 'padlen' becomes superfluous, so remove it. [ bp: - Adjust commit message - remove a stale comment about needing to pad - add a comment in optimize_nops() - exit early if the NOP verif. loop catches a mismatch - function should not not add NOPs in that case - fix the "optimized NOPs" offsets output ] Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov <bp@suse.de> Signed-off-by: Ingo Molnar <mingo@kernel.org> Link: https://lkml.kernel.org/r/20210326151259.442992235@infradead.org
2021-04-02Merge branch 'x86/cpu' into WIP.x86/core, to merge the NOP changes & resolve ↵Ingo Molnar1-5/+8
a semantic conflict Conflict-merge this main commit in essence: a89dfde3dc3c: ("x86: Remove dynamic NOP selection") With this upstream commit: b90829704780: ("bpf: Use NOP_ATOMIC5 instead of emit_nops(&prog, 5) for BPF_TRAMP_F_CALL_ORIG") Semantic merge conflict: arch/x86/net/bpf_jit_comp.c - memcpy(prog, ideal_nops[NOP_ATOMIC5], X86_PATCH_SIZE); + memcpy(prog, x86_nops[5], X86_PATCH_SIZE); Signed-off-by: Ingo Molnar <mingo@kernel.org>
2021-03-15objtool/x86: Use asm/nops.hPeter Zijlstra1-5/+8
Since the kernel will rely on a single canonical set of NOPs, make sure objtool uses the exact same ones. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Borislav Petkov <bp@suse.de> Link: https://lkml.kernel.org/r/20210312115749.136357911@infradead.org
2021-03-15tools/objtool: Convert to insn_decode()Borislav Petkov1-5/+4
Simplify code, no functional changes. Signed-off-by: Borislav Petkov <bp@suse.de> Link: https://lkml.kernel.org/r/20210304174237.31945-18-bp@alien8.de
2021-03-06objtool,x86: More ModRM sugarPeter Zijlstra1-11/+17
Better helpers to decode ModRM. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lkml.kernel.org/r/YCZB/ljatFXqQbm8@hirez.programming.kicks-ass.net
2021-03-06objtool,x86: Rewrite ADD/SUB/ANDPeter Zijlstra1-19/+51
Support sign extending and imm8 forms. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Link: https://lkml.kernel.org/r/20210211173627.588366777@infradead.org
2021-03-06objtool,x86: Support %riz encodingsPeter Zijlstra1-19/+48
When there's a SIB byte, the register otherwise denoted by r/m will then be denoted by SIB.base REX.b will now extend this. SIB.index == SP is magic and notes an index value zero. This means that there's a bunch of alternative (longer) encodings for the same thing. Eg. 'ModRM.mod != 3, ModRM.r/m = AX' can be encoded as 'ModRM.mod != 3, ModRM.r/m = SP, SIB.base = AX, SIB.index = SP' which is actually 4 different encodings because the value of SIB.scale is irrelevant, giving rise to 5 different but equal encodings. Support these encodings and clean up the SIB handling in general. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Link: https://lkml.kernel.org/r/20210211173627.472967498@infradead.org
2021-03-06objtool,x86: Simplify register decodePeter Zijlstra1-40/+39
Since the CFI_reg number now matches the instruction encoding order do away with the op_to_cfi_reg[] and use direct assignment. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Link: https://lkml.kernel.org/r/20210211173627.362004522@infradead.org
2021-03-06objtool,x86: Rewrite LEAVEPeter Zijlstra1-3/+11
Since we can now have multiple stack-ops per instruction, we don't need to special case LEAVE and can simply emit the composite operations. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Link: https://lkml.kernel.org/r/20210211173627.253273977@infradead.org
2021-03-06objtool,x86: Rewrite LEA decodePeter Zijlstra1-58/+28
Current LEA decoding is a bunch of special cases, properly decode the instruction, with exception of full SIB and RIP-relative modes. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Link: https://lkml.kernel.org/r/20210211173627.143250641@infradead.org
2021-03-06objtool,x86: Renumber CFI_regPeter Zijlstra1-6/+6
Make them match the instruction encoding numbering. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Link: https://lkml.kernel.org/r/20210211173627.033720313@infradead.org
2021-02-10objtool,x86: Additionally decode: mov %rsp, (%reg)Peter Zijlstra1-8/+34
Where we already decode: mov %rsp, %reg, also decode mov %rsp, (%reg). Nothing should match for this new stack-op. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com>
2021-01-26objtool: Combine UNWIND_HINT_RET_OFFSET and UNWIND_HINT_FUNCJosh Poimboeuf1-2/+2
The ORC metadata generated for UNWIND_HINT_FUNC isn't actually very func-like. With certain usages it can cause stack state mismatches because it doesn't set the return address (CFI_RA). Also, users of UNWIND_HINT_RET_OFFSET no longer need to set a custom return stack offset. Instead they just need to specify a func-like situation, so the current ret_offset code is hacky for no good reason. Solve both problems by simplifying the RET_OFFSET handling and converting it into a more useful UNWIND_HINT_FUNC. If we end up needing the old 'ret_offset' functionality again in the future, we should be able to support it pretty easily with the addition of a custom 'sp_offset' in UNWIND_HINT_FUNC. Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lore.kernel.org/r/db9d1f5d79dddfbb3725ef6d8ec3477ad199948d.1611263462.git.jpoimboe@redhat.com
2021-01-26objtool: Fix retpoline detection in asm codeJosh Poimboeuf1-1/+1
The JMP_NOSPEC macro branches to __x86_retpoline_*() rather than the __x86_indirect_thunk_*() wrappers used by C code. Detect jumps to __x86_retpoline_*() as retpoline dynamic jumps. Presumably this doesn't trigger a user-visible bug. I only found it when testing vmlinux.o validation. Fixes: 39b735332cb8 ("objtool: Detect jumps to retpoline thunks") Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lore.kernel.org/r/31f5833e2e4f01e3d755889ac77e3661e906c09f.1611263461.git.jpoimboe@redhat.com
generated by cgit v1.2.3 (git 2.39.0) at 2024-08-08 16:15:19 +0300