* [musl] Can’t build musl with lto=thin @ 2021-01-29 12:19 Jiahao XU 2021-01-29 20:04 ` Szabolcs Nagy 2021-01-30 20:12 ` Rich Felker 0 siblings, 2 replies; 21+ messages in thread From: Jiahao XU @ 2021-01-29 12:19 UTC (permalink / raw) To: musl [-- Attachment #1: Type: text/plain, Size: 542 bytes --] musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: ld.lld: error: undefined hidden symbol: __dls2 >>> referenced by ld-temp.o >>> lto.tmp:(_dlstart_c) >>> did you mean: __dls3 >>> defined in: lto.tmp I am using CFLAGS=‘-march=native -mtune=native -Oz -flto -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. No configure option is supplied. Get Outlook for iOS<https://aka.ms/o0ukef> [-- Attachment #2: Type: text/html, Size: 1169 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-29 12:19 [musl] Can’t build musl with lto=thin Jiahao XU @ 2021-01-29 20:04 ` Szabolcs Nagy 2021-01-29 20:53 ` Fangrui Song 2021-01-30 20:12 ` Rich Felker 1 sibling, 1 reply; 21+ messages in thread From: Szabolcs Nagy @ 2021-01-29 20:04 UTC (permalink / raw) To: Jiahao XU; +Cc: musl * Jiahao XU <Jiahao_XU@outlook.com> [2021-01-29 12:19:42 +0000]: > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > ld.lld: error: undefined hidden symbol: __dls2 > >>> referenced by ld-temp.o > >>> lto.tmp:(_dlstart_c) > >>> did you mean: __dls3 > >>> defined in: lto.tmp looks like a clang bug if -flto changes behaviour, doesn't it? > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > No configure option is supplied. > > Get Outlook for iOS<https://aka.ms/o0ukef> ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-29 20:04 ` Szabolcs Nagy @ 2021-01-29 20:53 ` Fangrui Song 0 siblings, 0 replies; 21+ messages in thread From: Fangrui Song @ 2021-01-29 20:53 UTC (permalink / raw) To: Jiahao XU; +Cc: musl, Szabolcs Nagy On Fri, Jan 29, 2021 at 12:04 PM Szabolcs Nagy <nsz@port70.net> wrote: > > * Jiahao XU <Jiahao_XU@outlook.com> [2021-01-29 12:19:42 +0000]: > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > ld.lld: error: undefined hidden symbol: __dls2 > > >>> referenced by ld-temp.o > > >>> lto.tmp:(_dlstart_c) > > >>> did you mean: __dls3 > > >>> defined in: lto.tmp > > > looks like a clang bug if -flto changes behaviour, doesn't it? https://www.openwall.com/lists/musl/2020/12/27/25 Clang tracks undefined symbols in module-level inline asm but does not know undefined symbols in function-scope inline asm. I don't know whether it makes sense to track function-scope inline asm - it requires to scan every instruction in the bitcode file which is a heavy operation and it is only relevant in this narrow usage. If you build this file with LTO, the LTO backend will internalize __dls2 and cause a subsequent "undefined reference" error unless you specify -u __dls2 to mark __dls2 as "visibible to a regular object file" (a workaround) > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > No configure option is supplied. > > > > Get Outlook for iOS<https://aka.ms/o0ukef> ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-29 12:19 [musl] Can’t build musl with lto=thin Jiahao XU 2021-01-29 20:04 ` Szabolcs Nagy @ 2021-01-30 20:12 ` Rich Felker 2021-01-30 20:59 ` Леонид Юрьев (Leonid Yuriev) [not found] ` <SYBP282MB020277E08EB49C9AC9BF90888AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM> 1 sibling, 2 replies; 21+ messages in thread From: Rich Felker @ 2021-01-30 20:12 UTC (permalink / raw) To: Jiahao XU; +Cc: musl On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > ld.lld: error: undefined hidden symbol: __dls2 > >>> referenced by ld-temp.o > >>> lto.tmp:(_dlstart_c) > >>> did you mean: __dls3 > >>> defined in: lto.tmp > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto ^^^^^^^^^^^^^^^^^^^^^ The -fmerge-all-constants option gives non-conforming language semantics and should not be used, but that's a separate issue. > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > No configure option is supplied. Otherwise, it's a known issue that LTO misses references from asm (both top-level and in functions). I think dlstart.lo and a few other files should just be built with LTO disabled; any LTO-type optimization in code that runs at this stage is inherently invalid, anyway. So something like (in config.mak): obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto Rich ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-30 20:12 ` Rich Felker @ 2021-01-30 20:59 ` Леонид Юрьев (Leonid Yuriev) [not found] ` <SYBP282MB020277E08EB49C9AC9BF90888AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM> 1 sibling, 0 replies; 21+ messages in thread From: Леонид Юрьев (Leonid Yuriev) @ 2021-01-30 20:59 UTC (permalink / raw) To: musl; +Cc: Jiahao XU I have some experience in solving problems with LTO, including for embedded projects. Therefore, perhaps my thoughts and advice will be useful. In general, LTO almost always requires a little refinement of source code but problems arise only with really dirty code or due to ill-conceived modularity. For instance, the ODR must not be violated. On Sat, Jan 30, 2021 at 11:12 PM Rich Felker <dalias@libc.org> wrote: > The -fmerge-all-constants option gives non-conforming language > semantics and should not be used, but that's a separate issue. There are no problems with constants, unless it's addresses are used to distinguish ones (and even more so to change). But this is easily workarounded by placing such constants inside an instance of a single static structure, etc. > Otherwise, it's a known issue that LTO misses references from asm > (both top-level and in functions). I think dlstart.lo and a few other > files should just be built with LTO disabled; any LTO-type > optimization in code that runs at this stage is inherently invalid, > anyway. So something like (in config.mak): This is easily workarounded by adding __attribute__((__used__)) or __attribute__((__externally_visible__)) to the function/variable definition, but requires binutils version >= 2.36. Regards, Leonid. ^ permalink raw reply [flat|nested] 21+ messages in thread
[parent not found: <SYBP282MB020277E08EB49C9AC9BF90888AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM>]
* Re: [musl] Can’t build musl with lto=thin [not found] ` <SYBP282MB020277E08EB49C9AC9BF90888AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM> @ 2021-01-30 23:30 ` Rich Felker [not found] ` <SYBP282MB0202BF16DD08A2F1537FD5428AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM> 0 siblings, 1 reply; 21+ messages in thread From: Rich Felker @ 2021-01-30 23:30 UTC (permalink / raw) To: Jiahao XU; +Cc: musl On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > So something like (in config.mak): > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > However, I still wasn’t able to statically linked with libc. > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. It's libc.a, not libc.so, that will be involved in making a static-linked binary. It's hard to know what's going wrong without more information. Can you run under a debugger and provide a backtrace, disassembly, and register dump for where the crash occurs? Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 7:12:31 AM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > ld.lld: error: undefined hidden symbol: __dls2 > > >>> referenced by ld-temp.o > > >>> lto.tmp:(_dlstart_c) > > >>> did you mean: __dls3 > > >>> defined in: lto.tmp > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > ^^^^^^^^^^^^^^^^^^^^^ > > The -fmerge-all-constants option gives non-conforming language > semantics and should not be used, but that's a separate issue. > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > No configure option is supplied. > > Otherwise, it's a known issue that LTO misses references from asm > (both top-level and in functions). I think dlstart.lo and a few other > files should just be built with LTO disabled; any LTO-type > optimization in code that runs at this stage is inherently invalid, > anyway. So something like (in config.mak): > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > Rich ^ permalink raw reply [flat|nested] 21+ messages in thread
[parent not found: <SYBP282MB0202BF16DD08A2F1537FD5428AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM>]
* Re: [musl] Can’t build musl with lto=thin [not found] ` <SYBP282MB0202BF16DD08A2F1537FD5428AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM> @ 2021-01-31 5:01 ` Rich Felker 2021-01-31 5:32 ` Jiahao XU 0 siblings, 1 reply; 21+ messages in thread From: Rich Felker @ 2021-01-31 5:01 UTC (permalink / raw) To: Jiahao XU; +Cc: musl On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > (gdb) bt > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > #2 0x0000000000000000 in ?? () This is not a static-linked program. decode_dyn is part of the dynamic linker. It looks to me like you've created some sort of weird hybrid executable that's not valid. Can you show the command lines you used to produce it? Also please keep list on CC when replying. Rich > (gdb) info r > > rax 0x1 1 > > rbx 0x7ffff7ffe2d8 140737354130136 > > rcx 0x5000 20480 > > rdx 0x200238 2097720 > > rsi 0x7fffffffd8d0 140737488345296 > > rdi 0x0 0 > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > r8 0x0 0 > > r9 0xfffffffffffff000 -4096 > > r10 0x800000 8388608 > > r11 0x200000 2097152 > > r12 0x7fffffffdcb8 140737488346296 > > r13 0x0 0 > > r14 0x7fffffffd8d0 140737488345296 > > r15 0x7fffffffdcb8 140737488346296 > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > eflags 0x10246 [ PF ZF IF RF ] > > cs 0x33 51 > > ss 0x2b 43 > > ds 0x0 0 > > es 0x0 0 > > fs 0x0 0 > > gs 0x0 0 > > > Disassembly of decode_dyn: > > > 0x00007ffff7ff58af <+0>: push %r14 > > 0x00007ffff7ff58b1 <+2>: push %rbx > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > Disassembly of decode_vec: > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > 0x00007ffff7ff5490 <+13>: inc %rax > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > 0x00007ffff7ff5495 <+18>: push $0x1 > > 0x00007ffff7ff5497 <+20>: pop %rax > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > 0x00007ffff7ff54c3 <+64>: ret > > > Jiahao XU > > Get Outlook for iOS<https://aka.ms/o0ukef> > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 10:30:12 AM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > However, I still wasn’t able to statically linked with libc. > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > It's libc.a, not libc.so, that will be involved in making a > static-linked binary. It's hard to know what's going wrong without > more information. Can you run under a debugger and provide a > backtrace, disassembly, and register dump for where the crash occurs? > > Rich > > > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 7:12:31 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > >>> referenced by ld-temp.o > > > >>> lto.tmp:(_dlstart_c) > > > >>> did you mean: __dls3 > > > >>> defined in: lto.tmp > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > ^^^^^^^^^^^^^^^^^^^^^ > > > > The -fmerge-all-constants option gives non-conforming language > > semantics and should not be used, but that's a separate issue. > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > No configure option is supplied. > > > > Otherwise, it's a known issue that LTO misses references from asm > > (both top-level and in functions). I think dlstart.lo and a few other > > files should just be built with LTO disabled; any LTO-type > > optimization in code that runs at this stage is inherently invalid, > > anyway. So something like (in config.mak): > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > Rich ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-31 5:01 ` Rich Felker @ 2021-01-31 5:32 ` Jiahao XU 2021-01-31 21:01 ` Rich Felker 0 siblings, 1 reply; 21+ messages in thread From: Jiahao XU @ 2021-01-31 5:32 UTC (permalink / raw) To: Rich Felker; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 7010 bytes --] I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as the generated `ld.musl-clang`, except for the last line, which I modified it to: exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Get Outlook for iOS<https://aka.ms/o0ukef> ________________________________ From: Rich Felker <dalias@libc.org> Sent: Sunday, January 31, 2021 4:01:22 PM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > (gdb) bt > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > #2 0x0000000000000000 in ?? () This is not a static-linked program. decode_dyn is part of the dynamic linker. It looks to me like you've created some sort of weird hybrid executable that's not valid. Can you show the command lines you used to produce it? Also please keep list on CC when replying. Rich > (gdb) info r > > rax 0x1 1 > > rbx 0x7ffff7ffe2d8 140737354130136 > > rcx 0x5000 20480 > > rdx 0x200238 2097720 > > rsi 0x7fffffffd8d0 140737488345296 > > rdi 0x0 0 > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > r8 0x0 0 > > r9 0xfffffffffffff000 -4096 > > r10 0x800000 8388608 > > r11 0x200000 2097152 > > r12 0x7fffffffdcb8 140737488346296 > > r13 0x0 0 > > r14 0x7fffffffd8d0 140737488345296 > > r15 0x7fffffffdcb8 140737488346296 > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > eflags 0x10246 [ PF ZF IF RF ] > > cs 0x33 51 > > ss 0x2b 43 > > ds 0x0 0 > > es 0x0 0 > > fs 0x0 0 > > gs 0x0 0 > > > Disassembly of decode_dyn: > > > 0x00007ffff7ff58af <+0>: push %r14 > > 0x00007ffff7ff58b1 <+2>: push %rbx > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > Disassembly of decode_vec: > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > 0x00007ffff7ff5490 <+13>: inc %rax > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > 0x00007ffff7ff5495 <+18>: push $0x1 > > 0x00007ffff7ff5497 <+20>: pop %rax > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > 0x00007ffff7ff54c3 <+64>: ret > > > Jiahao XU > > Get Outlook for iOS<https://aka.ms/o0ukef> > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 10:30:12 AM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > However, I still wasn’t able to statically linked with libc. > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > It's libc.a, not libc.so, that will be involved in making a > static-linked binary. It's hard to know what's going wrong without > more information. Can you run under a debugger and provide a > backtrace, disassembly, and register dump for where the crash occurs? > > Rich > > > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 7:12:31 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > >>> referenced by ld-temp.o > > > >>> lto.tmp:(_dlstart_c) > > > >>> did you mean: __dls3 > > > >>> defined in: lto.tmp > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > ^^^^^^^^^^^^^^^^^^^^^ > > > > The -fmerge-all-constants option gives non-conforming language > > semantics and should not be used, but that's a separate issue. > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > No configure option is supplied. > > > > Otherwise, it's a known issue that LTO misses references from asm > > (both top-level and in functions). I think dlstart.lo and a few other > > files should just be built with LTO disabled; any LTO-type > > optimization in code that runs at this stage is inherently invalid, > > anyway. So something like (in config.mak): > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > Rich [-- Attachment #2: Type: text/html, Size: 14338 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-31 5:32 ` Jiahao XU @ 2021-01-31 21:01 ` Rich Felker 2021-01-31 22:29 ` Jiahao XU 2021-02-01 0:22 ` Jiahao XU 0 siblings, 2 replies; 21+ messages in thread From: Rich Felker @ 2021-01-31 21:01 UTC (permalink / raw) To: Jiahao XU; +Cc: musl On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Where is -static? Normally it does *not* work to add -static just to the ld command line. The compiler driver has to know that it's requesting static linking because it will pass a different command line to the linker based on that. > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > the generated `ld.musl-clang`, except for the last line, which I > modified it to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Try moving -static out from here (i.e. using the script unmodified except for requesting ld.lld) and see if that works. Note that a correctly linked executable will not have any INTERP in readelf -a output, so as long as you see INTERP anywhere there you're doing something wrong. Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 4:01:22 PM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > (gdb) bt > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > #2 0x0000000000000000 in ?? () > > This is not a static-linked program. decode_dyn is part of the dynamic > linker. It looks to me like you've created some sort of weird hybrid > executable that's not valid. Can you show the command lines you used > to produce it? > > Also please keep list on CC when replying. > > Rich > > > > (gdb) info r > > > > rax 0x1 1 > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > rcx 0x5000 20480 > > > > rdx 0x200238 2097720 > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > rdi 0x0 0 > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > r8 0x0 0 > > > > r9 0xfffffffffffff000 -4096 > > > > r10 0x800000 8388608 > > > > r11 0x200000 2097152 > > > > r12 0x7fffffffdcb8 140737488346296 > > > > r13 0x0 0 > > > > r14 0x7fffffffd8d0 140737488345296 > > > > r15 0x7fffffffdcb8 140737488346296 > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > cs 0x33 51 > > > > ss 0x2b 43 > > > > ds 0x0 0 > > > > es 0x0 0 > > > > fs 0x0 0 > > > > gs 0x0 0 > > > > > > Disassembly of decode_dyn: > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > Disassembly of decode_vec: > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > Jiahao XU > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 10:30:12 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > It's libc.a, not libc.so, that will be involved in making a > > static-linked binary. It's hard to know what's going wrong without > > more information. Can you run under a debugger and provide a > > backtrace, disassembly, and register dump for where the crash occurs? > > > > Rich > > > > > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > >>> referenced by ld-temp.o > > > > >>> lto.tmp:(_dlstart_c) > > > > >>> did you mean: __dls3 > > > > >>> defined in: lto.tmp > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > The -fmerge-all-constants option gives non-conforming language > > > semantics and should not be used, but that's a separate issue. > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > No configure option is supplied. > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > (both top-level and in functions). I think dlstart.lo and a few other > > > files should just be built with LTO disabled; any LTO-type > > > optimization in code that runs at this stage is inherently invalid, > > > anyway. So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Rich ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-31 21:01 ` Rich Felker @ 2021-01-31 22:29 ` Jiahao XU 2021-02-01 0:22 ` Jiahao XU 1 sibling, 0 replies; 21+ messages in thread From: Jiahao XU @ 2021-01-31 22:29 UTC (permalink / raw) To: Rich Felker; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 8461 bytes --] I tried to run “musl-clang -Oz -flto -s -static -fuse-ld=musl-clang-lld -Wl,—plugin-opt=O3,-O3 hello.c”, but it still doesn’t work. Note that I have removed -static from ld.musl-clang-lld Jiahao XU ________________________________ From: Rich Felker <dalias@libc.org> Sent: Monday, February 1, 2021 8:01:06 AM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Where is -static? Normally it does *not* work to add -static just to the ld command line. The compiler driver has to know that it's requesting static linking because it will pass a different command line to the linker based on that. > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > the generated `ld.musl-clang`, except for the last line, which I > modified it to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Try moving -static out from here (i.e. using the script unmodified except for requesting ld.lld) and see if that works. Note that a correctly linked executable will not have any INTERP in readelf -a output, so as long as you see INTERP anywhere there you're doing something wrong. Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 4:01:22 PM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > (gdb) bt > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > #2 0x0000000000000000 in ?? () > > This is not a static-linked program. decode_dyn is part of the dynamic > linker. It looks to me like you've created some sort of weird hybrid > executable that's not valid. Can you show the command lines you used > to produce it? > > Also please keep list on CC when replying. > > Rich > > > > (gdb) info r > > > > rax 0x1 1 > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > rcx 0x5000 20480 > > > > rdx 0x200238 2097720 > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > rdi 0x0 0 > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > r8 0x0 0 > > > > r9 0xfffffffffffff000 -4096 > > > > r10 0x800000 8388608 > > > > r11 0x200000 2097152 > > > > r12 0x7fffffffdcb8 140737488346296 > > > > r13 0x0 0 > > > > r14 0x7fffffffd8d0 140737488345296 > > > > r15 0x7fffffffdcb8 140737488346296 > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > cs 0x33 51 > > > > ss 0x2b 43 > > > > ds 0x0 0 > > > > es 0x0 0 > > > > fs 0x0 0 > > > > gs 0x0 0 > > > > > > Disassembly of decode_dyn: > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > Disassembly of decode_vec: > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > Jiahao XU > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 10:30:12 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > It's libc.a, not libc.so, that will be involved in making a > > static-linked binary. It's hard to know what's going wrong without > > more information. Can you run under a debugger and provide a > > backtrace, disassembly, and register dump for where the crash occurs? > > > > Rich > > > > > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > >>> referenced by ld-temp.o > > > > >>> lto.tmp:(_dlstart_c) > > > > >>> did you mean: __dls3 > > > > >>> defined in: lto.tmp > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > The -fmerge-all-constants option gives non-conforming language > > > semantics and should not be used, but that's a separate issue. > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > No configure option is supplied. > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > (both top-level and in functions). I think dlstart.lo and a few other > > > files should just be built with LTO disabled; any LTO-type > > > optimization in code that runs at this stage is inherently invalid, > > > anyway. So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Rich [-- Attachment #2: Type: text/html, Size: 16449 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-01-31 21:01 ` Rich Felker 2021-01-31 22:29 ` Jiahao XU @ 2021-02-01 0:22 ` Jiahao XU 2021-02-01 0:31 ` Jiahao XU 1 sibling, 1 reply; 21+ messages in thread From: Jiahao XU @ 2021-02-01 0:22 UTC (permalink / raw) To: Rich Felker; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 8520 bytes --] I have finally succeded to produce a statically linked hello_world executable. I modified the last line of musl-clang-lld to: exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -l:libc.a —no-dynamic-linker and everything works fine now. Jiahao XU ________________________________ From: Rich Felker <dalias@libc.org> Sent: Monday, February 1, 2021 8:01:06 AM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Where is -static? Normally it does *not* work to add -static just to the ld command line. The compiler driver has to know that it's requesting static linking because it will pass a different command line to the linker based on that. > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > the generated `ld.musl-clang`, except for the last line, which I > modified it to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Try moving -static out from here (i.e. using the script unmodified except for requesting ld.lld) and see if that works. Note that a correctly linked executable will not have any INTERP in readelf -a output, so as long as you see INTERP anywhere there you're doing something wrong. Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 4:01:22 PM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > (gdb) bt > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > #2 0x0000000000000000 in ?? () > > This is not a static-linked program. decode_dyn is part of the dynamic > linker. It looks to me like you've created some sort of weird hybrid > executable that's not valid. Can you show the command lines you used > to produce it? > > Also please keep list on CC when replying. > > Rich > > > > (gdb) info r > > > > rax 0x1 1 > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > rcx 0x5000 20480 > > > > rdx 0x200238 2097720 > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > rdi 0x0 0 > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > r8 0x0 0 > > > > r9 0xfffffffffffff000 -4096 > > > > r10 0x800000 8388608 > > > > r11 0x200000 2097152 > > > > r12 0x7fffffffdcb8 140737488346296 > > > > r13 0x0 0 > > > > r14 0x7fffffffd8d0 140737488345296 > > > > r15 0x7fffffffdcb8 140737488346296 > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > cs 0x33 51 > > > > ss 0x2b 43 > > > > ds 0x0 0 > > > > es 0x0 0 > > > > fs 0x0 0 > > > > gs 0x0 0 > > > > > > Disassembly of decode_dyn: > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > Disassembly of decode_vec: > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > Jiahao XU > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 10:30:12 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > It's libc.a, not libc.so, that will be involved in making a > > static-linked binary. It's hard to know what's going wrong without > > more information. Can you run under a debugger and provide a > > backtrace, disassembly, and register dump for where the crash occurs? > > > > Rich > > > > > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > >>> referenced by ld-temp.o > > > > >>> lto.tmp:(_dlstart_c) > > > > >>> did you mean: __dls3 > > > > >>> defined in: lto.tmp > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > The -fmerge-all-constants option gives non-conforming language > > > semantics and should not be used, but that's a separate issue. > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > No configure option is supplied. > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > (both top-level and in functions). I think dlstart.lo and a few other > > > files should just be built with LTO disabled; any LTO-type > > > optimization in code that runs at this stage is inherently invalid, > > > anyway. So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Rich [-- Attachment #2: Type: text/html, Size: 16897 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 0:22 ` Jiahao XU @ 2021-02-01 0:31 ` Jiahao XU 2021-02-01 0:44 ` Jiahao XU 2021-02-01 17:53 ` Rich Felker 0 siblings, 2 replies; 21+ messages in thread From: Jiahao XU @ 2021-02-01 0:31 UTC (permalink / raw) To: Rich Felker; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 9711 bytes --] Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. .data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. $ bloaty gc-section-a.out -- no-gc-section.a.out FILE SIZE VM SIZE -------------- -------------- +18% +56 -3.2% -8 .data [NEW] +6 [NEW] +6 [LOAD #2 [RX]] [DEL] -4 -66.7% -8 [LOAD #4 [RW]] -72.7% -8 [ = ] 0 [Unmapped] -32.0% -64 [ = ] 0 .comment -99.4% -19.4Ki -99.7% -19.4Ki .rodata -88.0% -22.3Ki -88.2% -22.3Ki .text -89.4% -41.8Ki -88.5% -41.8Ki TOTAL Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:22:21 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin I have finally succeded to produce a statically linked hello_world executable. I modified the last line of musl-clang-lld to: exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -l:libc.a —no-dynamic-linker and everything works fine now. Jiahao XU ________________________________ From: Rich Felker <dalias@libc.org> Sent: Monday, February 1, 2021 8:01:06 AM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Where is -static? Normally it does *not* work to add -static just to the ld command line. The compiler driver has to know that it's requesting static linking because it will pass a different command line to the linker based on that. > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > the generated `ld.musl-clang`, except for the last line, which I > modified it to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Try moving -static out from here (i.e. using the script unmodified except for requesting ld.lld) and see if that works. Note that a correctly linked executable will not have any INTERP in readelf -a output, so as long as you see INTERP anywhere there you're doing something wrong. Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 4:01:22 PM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > (gdb) bt > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > #2 0x0000000000000000 in ?? () > > This is not a static-linked program. decode_dyn is part of the dynamic > linker. It looks to me like you've created some sort of weird hybrid > executable that's not valid. Can you show the command lines you used > to produce it? > > Also please keep list on CC when replying. > > Rich > > > > (gdb) info r > > > > rax 0x1 1 > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > rcx 0x5000 20480 > > > > rdx 0x200238 2097720 > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > rdi 0x0 0 > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > r8 0x0 0 > > > > r9 0xfffffffffffff000 -4096 > > > > r10 0x800000 8388608 > > > > r11 0x200000 2097152 > > > > r12 0x7fffffffdcb8 140737488346296 > > > > r13 0x0 0 > > > > r14 0x7fffffffd8d0 140737488345296 > > > > r15 0x7fffffffdcb8 140737488346296 > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > cs 0x33 51 > > > > ss 0x2b 43 > > > > ds 0x0 0 > > > > es 0x0 0 > > > > fs 0x0 0 > > > > gs 0x0 0 > > > > > > Disassembly of decode_dyn: > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > Disassembly of decode_vec: > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > Jiahao XU > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 10:30:12 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > It's libc.a, not libc.so, that will be involved in making a > > static-linked binary. It's hard to know what's going wrong without > > more information. Can you run under a debugger and provide a > > backtrace, disassembly, and register dump for where the crash occurs? > > > > Rich > > > > > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > >>> referenced by ld-temp.o > > > > >>> lto.tmp:(_dlstart_c) > > > > >>> did you mean: __dls3 > > > > >>> defined in: lto.tmp > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > The -fmerge-all-constants option gives non-conforming language > > > semantics and should not be used, but that's a separate issue. > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > No configure option is supplied. > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > (both top-level and in functions). I think dlstart.lo and a few other > > > files should just be built with LTO disabled; any LTO-type > > > optimization in code that runs at this stage is inherently invalid, > > > anyway. So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Rich [-- Attachment #2: Type: text/html, Size: 22557 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 0:31 ` Jiahao XU @ 2021-02-01 0:44 ` Jiahao XU 2021-02-01 0:50 ` Jiahao XU 2021-02-01 17:53 ` Rich Felker 1 sibling, 1 reply; 21+ messages in thread From: Jiahao XU @ 2021-02-01 0:44 UTC (permalink / raw) To: Rich Felker; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 10082 bytes --] Correction: The executable is cut from 48KB to 8KB. I mixed up the numbers from du and ls. Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:31:47 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. .data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. $ bloaty gc-section-a.out -- no-gc-section.a.out FILE SIZE VM SIZE -------------- -------------- +18% +56 -3.2% -8 .data [NEW] +6 [NEW] +6 [LOAD #2 [RX]] [DEL] -4 -66.7% -8 [LOAD #4 [RW]] -72.7% -8 [ = ] 0 [Unmapped] -32.0% -64 [ = ] 0 .comment -99.4% -19.4Ki -99.7% -19.4Ki .rodata -88.0% -22.3Ki -88.2% -22.3Ki .text -89.4% -41.8Ki -88.5% -41.8Ki TOTAL Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:22:21 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin I have finally succeded to produce a statically linked hello_world executable. I modified the last line of musl-clang-lld to: exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -l:libc.a —no-dynamic-linker and everything works fine now. Jiahao XU ________________________________ From: Rich Felker <dalias@libc.org> Sent: Monday, February 1, 2021 8:01:06 AM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Where is -static? Normally it does *not* work to add -static just to the ld command line. The compiler driver has to know that it's requesting static linking because it will pass a different command line to the linker based on that. > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > the generated `ld.musl-clang`, except for the last line, which I > modified it to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Try moving -static out from here (i.e. using the script unmodified except for requesting ld.lld) and see if that works. Note that a correctly linked executable will not have any INTERP in readelf -a output, so as long as you see INTERP anywhere there you're doing something wrong. Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 4:01:22 PM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > (gdb) bt > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > #2 0x0000000000000000 in ?? () > > This is not a static-linked program. decode_dyn is part of the dynamic > linker. It looks to me like you've created some sort of weird hybrid > executable that's not valid. Can you show the command lines you used > to produce it? > > Also please keep list on CC when replying. > > Rich > > > > (gdb) info r > > > > rax 0x1 1 > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > rcx 0x5000 20480 > > > > rdx 0x200238 2097720 > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > rdi 0x0 0 > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > r8 0x0 0 > > > > r9 0xfffffffffffff000 -4096 > > > > r10 0x800000 8388608 > > > > r11 0x200000 2097152 > > > > r12 0x7fffffffdcb8 140737488346296 > > > > r13 0x0 0 > > > > r14 0x7fffffffd8d0 140737488345296 > > > > r15 0x7fffffffdcb8 140737488346296 > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > cs 0x33 51 > > > > ss 0x2b 43 > > > > ds 0x0 0 > > > > es 0x0 0 > > > > fs 0x0 0 > > > > gs 0x0 0 > > > > > > Disassembly of decode_dyn: > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > Disassembly of decode_vec: > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > Jiahao XU > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 10:30:12 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > It's libc.a, not libc.so, that will be involved in making a > > static-linked binary. It's hard to know what's going wrong without > > more information. Can you run under a debugger and provide a > > backtrace, disassembly, and register dump for where the crash occurs? > > > > Rich > > > > > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > >>> referenced by ld-temp.o > > > > >>> lto.tmp:(_dlstart_c) > > > > >>> did you mean: __dls3 > > > > >>> defined in: lto.tmp > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > The -fmerge-all-constants option gives non-conforming language > > > semantics and should not be used, but that's a separate issue. > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > No configure option is supplied. > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > (both top-level and in functions). I think dlstart.lo and a few other > > > files should just be built with LTO disabled; any LTO-type > > > optimization in code that runs at this stage is inherently invalid, > > > anyway. So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Rich [-- Attachment #2: Type: text/html, Size: 23567 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 0:44 ` Jiahao XU @ 2021-02-01 0:50 ` Jiahao XU 2021-02-01 13:20 ` Jiahao XU 0 siblings, 1 reply; 21+ messages in thread From: Jiahao XU @ 2021-02-01 0:50 UTC (permalink / raw) To: Rich Felker; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 10535 bytes --] Correction: The right number should be the executable is cut from 47KB to 5.0KB. The previous numbers are retrieved using du -hs, which calculate disk usage, not file size. Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:44:49 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin Correction: The executable is cut from 48KB to 8KB. I mixed up the numbers from du and ls. Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:31:47 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. .data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. $ bloaty gc-section-a.out -- no-gc-section.a.out FILE SIZE VM SIZE -------------- -------------- +18% +56 -3.2% -8 .data [NEW] +6 [NEW] +6 [LOAD #2 [RX]] [DEL] -4 -66.7% -8 [LOAD #4 [RW]] -72.7% -8 [ = ] 0 [Unmapped] -32.0% -64 [ = ] 0 .comment -99.4% -19.4Ki -99.7% -19.4Ki .rodata -88.0% -22.3Ki -88.2% -22.3Ki .text -89.4% -41.8Ki -88.5% -41.8Ki TOTAL Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:22:21 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin I have finally succeded to produce a statically linked hello_world executable. I modified the last line of musl-clang-lld to: exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -l:libc.a —no-dynamic-linker and everything works fine now. Jiahao XU ________________________________ From: Rich Felker <dalias@libc.org> Sent: Monday, February 1, 2021 8:01:06 AM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Where is -static? Normally it does *not* work to add -static just to the ld command line. The compiler driver has to know that it's requesting static linking because it will pass a different command line to the linker based on that. > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > the generated `ld.musl-clang`, except for the last line, which I > modified it to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Try moving -static out from here (i.e. using the script unmodified except for requesting ld.lld) and see if that works. Note that a correctly linked executable will not have any INTERP in readelf -a output, so as long as you see INTERP anywhere there you're doing something wrong. Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 4:01:22 PM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > (gdb) bt > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > #2 0x0000000000000000 in ?? () > > This is not a static-linked program. decode_dyn is part of the dynamic > linker. It looks to me like you've created some sort of weird hybrid > executable that's not valid. Can you show the command lines you used > to produce it? > > Also please keep list on CC when replying. > > Rich > > > > (gdb) info r > > > > rax 0x1 1 > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > rcx 0x5000 20480 > > > > rdx 0x200238 2097720 > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > rdi 0x0 0 > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > r8 0x0 0 > > > > r9 0xfffffffffffff000 -4096 > > > > r10 0x800000 8388608 > > > > r11 0x200000 2097152 > > > > r12 0x7fffffffdcb8 140737488346296 > > > > r13 0x0 0 > > > > r14 0x7fffffffd8d0 140737488345296 > > > > r15 0x7fffffffdcb8 140737488346296 > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > cs 0x33 51 > > > > ss 0x2b 43 > > > > ds 0x0 0 > > > > es 0x0 0 > > > > fs 0x0 0 > > > > gs 0x0 0 > > > > > > Disassembly of decode_dyn: > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > Disassembly of decode_vec: > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > Jiahao XU > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 10:30:12 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > It's libc.a, not libc.so, that will be involved in making a > > static-linked binary. It's hard to know what's going wrong without > > more information. Can you run under a debugger and provide a > > backtrace, disassembly, and register dump for where the crash occurs? > > > > Rich > > > > > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > >>> referenced by ld-temp.o > > > > >>> lto.tmp:(_dlstart_c) > > > > >>> did you mean: __dls3 > > > > >>> defined in: lto.tmp > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > The -fmerge-all-constants option gives non-conforming language > > > semantics and should not be used, but that's a separate issue. > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > No configure option is supplied. > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > (both top-level and in functions). I think dlstart.lo and a few other > > > files should just be built with LTO disabled; any LTO-type > > > optimization in code that runs at this stage is inherently invalid, > > > anyway. So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Rich [-- Attachment #2: Type: text/html, Size: 24726 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 0:50 ` Jiahao XU @ 2021-02-01 13:20 ` Jiahao XU 2021-02-01 16:55 ` Nick Bray 0 siblings, 1 reply; 21+ messages in thread From: Jiahao XU @ 2021-02-01 13:20 UTC (permalink / raw) To: Rich Felker; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 10854 bytes --] Thank you very much for helping me, RIch. Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:50:04 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin Correction: The right number should be the executable is cut from 47KB to 5.0KB. The previous numbers are retrieved using du -hs, which calculate disk usage, not file size. Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:44:49 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin Correction: The executable is cut from 48KB to 8KB. I mixed up the numbers from du and ls. Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:31:47 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. .data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. $ bloaty gc-section-a.out -- no-gc-section.a.out FILE SIZE VM SIZE -------------- -------------- +18% +56 -3.2% -8 .data [NEW] +6 [NEW] +6 [LOAD #2 [RX]] [DEL] -4 -66.7% -8 [LOAD #4 [RW]] -72.7% -8 [ = ] 0 [Unmapped] -32.0% -64 [ = ] 0 .comment -99.4% -19.4Ki -99.7% -19.4Ki .rodata -88.0% -22.3Ki -88.2% -22.3Ki .text -89.4% -41.8Ki -88.5% -41.8Ki TOTAL Jiahao XU ________________________________ From: Jiahao XU <Jiahao_XU@outlook.com> Sent: Monday, February 1, 2021 11:22:21 AM To: Rich Felker <dalias@libc.org> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin I have finally succeded to produce a statically linked hello_world executable. I modified the last line of musl-clang-lld to: exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -l:libc.a —no-dynamic-linker and everything works fine now. Jiahao XU ________________________________ From: Rich Felker <dalias@libc.org> Sent: Monday, February 1, 2021 8:01:06 AM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. Where is -static? Normally it does *not* work to add -static just to the ld command line. The compiler driver has to know that it's requesting static linking because it will pass a different command line to the linker based on that. > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > the generated `ld.musl-clang`, except for the last line, which I > modified it to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” Try moving -static out from here (i.e. using the script unmodified except for requesting ld.lld) and see if that works. Note that a correctly linked executable will not have any INTERP in readelf -a output, so as long as you see INTERP anywhere there you're doing something wrong. Rich > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Sunday, January 31, 2021 4:01:22 PM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > (gdb) bt > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > #2 0x0000000000000000 in ?? () > > This is not a static-linked program. decode_dyn is part of the dynamic > linker. It looks to me like you've created some sort of weird hybrid > executable that's not valid. Can you show the command lines you used > to produce it? > > Also please keep list on CC when replying. > > Rich > > > > (gdb) info r > > > > rax 0x1 1 > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > rcx 0x5000 20480 > > > > rdx 0x200238 2097720 > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > rdi 0x0 0 > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > r8 0x0 0 > > > > r9 0xfffffffffffff000 -4096 > > > > r10 0x800000 8388608 > > > > r11 0x200000 2097152 > > > > r12 0x7fffffffdcb8 140737488346296 > > > > r13 0x0 0 > > > > r14 0x7fffffffd8d0 140737488345296 > > > > r15 0x7fffffffdcb8 140737488346296 > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > cs 0x33 51 > > > > ss 0x2b 43 > > > > ds 0x0 0 > > > > es 0x0 0 > > > > fs 0x0 0 > > > > gs 0x0 0 > > > > > > Disassembly of decode_dyn: > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > Disassembly of decode_vec: > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > Jiahao XU > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 10:30:12 AM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > It's libc.a, not libc.so, that will be involved in making a > > static-linked binary. It's hard to know what's going wrong without > > more information. Can you run under a debugger and provide a > > backtrace, disassembly, and register dump for where the crash occurs? > > > > Rich > > > > > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > >>> referenced by ld-temp.o > > > > >>> lto.tmp:(_dlstart_c) > > > > >>> did you mean: __dls3 > > > > >>> defined in: lto.tmp > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > The -fmerge-all-constants option gives non-conforming language > > > semantics and should not be used, but that's a separate issue. > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > No configure option is supplied. > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > (both top-level and in functions). I think dlstart.lo and a few other > > > files should just be built with LTO disabled; any LTO-type > > > optimization in code that runs at this stage is inherently invalid, > > > anyway. So something like (in config.mak): > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > Rich [-- Attachment #2: Type: text/html, Size: 25567 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 13:20 ` Jiahao XU @ 2021-02-01 16:55 ` Nick Bray 0 siblings, 0 replies; 21+ messages in thread From: Nick Bray @ 2021-02-01 16:55 UTC (permalink / raw) To: musl; +Cc: Rich Felker [-- Attachment #1: Type: text/plain, Size: 12489 bytes --] Small warning: When I statically linked in Musl and performed LTO on the entire executable with Clang, I ran into a few bugs. Specifically, when Clang could rewrite libc calls (memcmp => bcmp, for example) the rewritten call no longer had any calling convention information in the IR. If LTO then tried to inline the call, Clang got confused and emitted a UDF instead of an actual call. I was working on aarch64. I believe the loss of calling convention is architecture independent, but I don't know what the other backends do with the malformed IR. I was hesitant to mail the list because I _think_ there's a patch in flight and it may have landed, but I have been too busy to follow up. I figured I should at least mention the problems that I have seen, inlining across the libc boundary. On Mon, Feb 1, 2021 at 5:28 AM Jiahao XU <Jiahao_XU@outlook.com> wrote: > Thank you very much for helping me, RIch. > > Jiahao XU > ------------------------------ > *From:* Jiahao XU <Jiahao_XU@outlook.com> > *Sent:* Monday, February 1, 2021 11:50:04 AM > *To:* Rich Felker <dalias@libc.org> > *Cc:* musl@lists.openwall.com <musl@lists.openwall.com> > *Subject:* Re: [musl] Can’t build musl with lto=thin > > Correction: The right number should be the executable is cut from 47KB to > 5.0KB. > > The previous numbers are retrieved using du -hs, which calculate disk > usage, not file size. > > Jiahao XU > ------------------------------ > *From:* Jiahao XU <Jiahao_XU@outlook.com> > *Sent:* Monday, February 1, 2021 11:44:49 AM > *To:* Rich Felker <dalias@libc.org> > *Cc:* musl@lists.openwall.com <musl@lists.openwall.com> > *Subject:* Re: [musl] Can’t build musl with lto=thin > > Correction: The executable is cut from 48KB to 8KB. > > I mixed up the numbers from du and ls. > > Jiahao XU > ------------------------------ > *From:* Jiahao XU <Jiahao_XU@outlook.com> > *Sent:* Monday, February 1, 2021 11:31:47 AM > *To:* Rich Felker <dalias@libc.org> > *Cc:* musl@lists.openwall.com <musl@lists.openwall.com> > *Subject:* Re: [musl] Can’t build musl with lto=thin > > Interesting enough, I found —gc-section used along with -flto can cut the > size of final hello_world executable from 48KB to 5KB. > > After investigating with bloaty, I found that —gc-section along with -flto > is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from > 19.5 KiB to 120 bytes. > .data section however, seen an increase from 316 bytes to 372 bytes, but > the VM size is cut from 252 to 244 bytes. > > $ bloaty gc-section-a.out -- no-gc-section.a.out > > FILE SIZE VM SIZE > > -------------- -------------- > > +18% +56 -3.2% -8 .data > > [NEW] +6 [NEW] +6 [LOAD #2 [RX]] > > [DEL] -4 -66.7% -8 [LOAD #4 [RW]] > > -72.7% -8 [ = ] 0 [Unmapped] > > -32.0% -64 [ = ] 0 .comment > > -99.4% -19.4Ki -99.7% -19.4Ki .rodata > > -88.0% -22.3Ki -88.2% -22.3Ki .text > > -89.4% -41.8Ki -88.5% -41.8Ki TOTAL > > > Jiahao XU > ------------------------------ > *From:* Jiahao XU <Jiahao_XU@outlook.com> > *Sent:* Monday, February 1, 2021 11:22:21 AM > *To:* Rich Felker <dalias@libc.org> > *Cc:* musl@lists.openwall.com <musl@lists.openwall.com> > *Subject:* Re: [musl] Can’t build musl with lto=thin > > I have finally succeded to produce a statically linked hello_world > executable. > > I modified the last line of musl-clang-lld to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -l:libc.a > —no-dynamic-linker > > and everything works fine now. > > Jiahao XU > ------------------------------ > *From:* Rich Felker <dalias@libc.org> > *Sent:* Monday, February 1, 2021 8:01:06 AM > *To:* Jiahao XU <Jiahao_XU@outlook.com> > *Cc:* musl@lists.openwall.com <musl@lists.openwall.com> > *Subject:* Re: [musl] Can’t build musl with lto=thin > > On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static > -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. > > Where is -static? Normally it does *not* work to add -static just to > the ld command line. The compiler driver has to know that it's > requesting static linking because it will pass a different command > line to the linker based on that. > > > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > > the generated `ld.musl-clang`, except for the last line, which I > > modified it to: > > > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc > -dynamic-linker “$ldso” > > Try moving -static out from here (i.e. using the script unmodified > except for requesting ld.lld) and see if that works. Note that a > correctly linked executable will not have any INTERP in readelf -a > output, so as long as you see INTERP anywhere there you're doing > something wrong. > > Rich > > > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 4:01:22 PM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > > (gdb) bt > > > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > > > #2 0x0000000000000000 in ?? () > > > > This is not a static-linked program. decode_dyn is part of the dynamic > > linker. It looks to me like you've created some sort of weird hybrid > > executable that's not valid. Can you show the command lines you used > > to produce it? > > > > Also please keep list on CC when replying. > > > > Rich > > > > > > > (gdb) info r > > > > > > rax 0x1 1 > > > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > > > rcx 0x5000 20480 > > > > > > rdx 0x200238 2097720 > > > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > > > rdi 0x0 0 > > > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > > > r8 0x0 0 > > > > > > r9 0xfffffffffffff000 -4096 > > > > > > r10 0x800000 8388608 > > > > > > r11 0x200000 2097152 > > > > > > r12 0x7fffffffdcb8 140737488346296 > > > > > > r13 0x0 0 > > > > > > r14 0x7fffffffd8d0 140737488345296 > > > > > > r15 0x7fffffffdcb8 140737488346296 > > > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > > > cs 0x33 51 > > > > > > ss 0x2b 43 > > > > > > ds 0x0 0 > > > > > > es 0x0 0 > > > > > > fs 0x0 0 > > > > > > gs 0x0 0 > > > > > > > > > Disassembly of decode_dyn: > > > > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > > > > Disassembly of decode_vec: > > > > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > > > > Jiahao XU > > > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 10:30:12 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > > So something like (in config.mak): > > > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > > > Thanks, with this I was able to build libc.so successfully with > clang and created a 3.5 KB hello world program using clang and lld. > > > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed > with ‘Segmentation fault (core dumped)’. > > > > > > It's libc.a, not libc.so, that will be involved in making a > > > static-linked binary. It's hard to know what's going wrong without > > > more information. Can you run under a debugger and provide a > > > backtrace, disassembly, and register dump for where the crash occurs? > > > > > > Rich > > > > > > > > > > ________________________________ > > > > From: Rich Felker <dalias@libc.org> > > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at > the final linking stage: > > > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > > >>> referenced by ld-temp.o > > > > > >>> lto.tmp:(_dlstart_c) > > > > > >>> did you mean: __dls3 > > > > > >>> defined in: lto.tmp > > > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > > > The -fmerge-all-constants option gives non-conforming language > > > > semantics and should not be used, but that's a separate issue. > > > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > > > No configure option is supplied. > > > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > > (both top-level and in functions). I think dlstart.lo and a few other > > > > files should just be built with LTO disabled; any LTO-type > > > > optimization in code that runs at this stage is inherently invalid, > > > > anyway. So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > > > Rich > [-- Attachment #2: Type: text/html, Size: 23997 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 0:31 ` Jiahao XU 2021-02-01 0:44 ` Jiahao XU @ 2021-02-01 17:53 ` Rich Felker 2021-02-01 18:06 ` Fangrui Song 1 sibling, 1 reply; 21+ messages in thread From: Rich Felker @ 2021-02-01 17:53 UTC (permalink / raw) To: Jiahao XU; +Cc: musl On Mon, Feb 01, 2021 at 12:31:47AM +0000, Jiahao XU wrote: > Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. > > After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. > ..data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. > > > $ bloaty gc-section-a.out -- no-gc-section.a.out > > FILE SIZE VM SIZE > > -------------- -------------- > > +18% +56 -3.2% -8 .data > > [NEW] +6 [NEW] +6 [LOAD #2 [RX]] > > [DEL] -4 -66.7% -8 [LOAD #4 [RW]] > > -72.7% -8 [ = ] 0 [Unmapped] > > -32.0% -64 [ = ] 0 .comment > > -99.4% -19.4Ki -99.7% -19.4Ki .rodata > > -88.0% -22.3Ki -88.2% -22.3Ki .text > > -89.4% -41.8Ki -88.5% -41.8Ki TOTAL What is included in your hello world? Mine, static linked normally (no LTO) is around 4k of text and virtually no rodata. (This is with GCC; I'm not using clang.) If I compile with -fno-builtin so printf doesn't get transformed to puts, there's about 16k of text and 3k of rodata, but still nowhere near the ~20k you saw. Rich > ________________________________ > From: Jiahao XU <Jiahao_XU@outlook.com> > Sent: Monday, February 1, 2021 11:22:21 AM > To: Rich Felker <dalias@libc.org> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > I have finally succeded to produce a statically linked hello_world executable. > > I modified the last line of musl-clang-lld to: > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -l:libc.a —no-dynamic-linker > > and everything works fine now. > > Jiahao XU > ________________________________ > From: Rich Felker <dalias@libc.org> > Sent: Monday, February 1, 2021 8:01:06 AM > To: Jiahao XU <Jiahao_XU@outlook.com> > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > Subject: Re: [musl] Can’t build musl with lto=thin > > On Sun, Jan 31, 2021 at 05:32:45AM +0000, Jiahao XU wrote: > > I used `musl-clang -Oz -flto -s -fuse-ld=musl-clang-lld-static -Wl,—plugin-opt=O3,-O3 hello.c` to produce the executable. > > Where is -static? Normally it does *not* work to add -static just to > the ld command line. The compiler driver has to know that it's > requesting static linking because it will pass a different command > line to the linker based on that. > > > Content of `/usr/local/musl/bin/ld.musl-clang-lld-static` is same as > > the generated `ld.musl-clang`, except for the last line, which I > > modified it to: > > > > exec $($cc -print-prog-name=ld.lld) -nostdlib “$@“ -static -lc -dynamic-linker “$ldso” > > Try moving -static out from here (i.e. using the script unmodified > except for requesting ld.lld) and see if that works. Note that a > correctly linked executable will not have any INTERP in readelf -a > output, so as long as you see INTERP anywhere there you're doing > something wrong. > > Rich > > > > ________________________________ > > From: Rich Felker <dalias@libc.org> > > Sent: Sunday, January 31, 2021 4:01:22 PM > > To: Jiahao XU <Jiahao_XU@outlook.com> > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > On Sat, Jan 30, 2021 at 11:44:48PM +0000, Jiahao XU wrote: > > > (gdb) bt > > > > > > #0 0x00007ffff7ff5498 in decode_vec () from /lib/ld-musl-x86_64.so.1 > > > > > > #1 0x00007ffff7ff58cb in decode_dyn () from /lib/ld-musl-x86_64.so.1 > > > > > > #2 0x0000000000000000 in ?? () > > > > This is not a static-linked program. decode_dyn is part of the dynamic > > linker. It looks to me like you've created some sort of weird hybrid > > executable that's not valid. Can you show the command lines you used > > to produce it? > > > > Also please keep list on CC when replying. > > > > Rich > > > > > > > (gdb) info r > > > > > > rax 0x1 1 > > > > > > rbx 0x7ffff7ffe2d8 140737354130136 > > > > > > rcx 0x5000 20480 > > > > > > rdx 0x200238 2097720 > > > > > > rsi 0x7fffffffd8d0 140737488345296 > > > > > > rdi 0x0 0 > > > > > > rbp 0x7ffff7ffe2d8 0x7ffff7ffe2d8 <__dls3.app> > > > > > > rsp 0x7fffffffd8c8 0x7fffffffd8c8 > > > > > > r8 0x0 0 > > > > > > r9 0xfffffffffffff000 -4096 > > > > > > r10 0x800000 8388608 > > > > > > r11 0x200000 2097152 > > > > > > r12 0x7fffffffdcb8 140737488346296 > > > > > > r13 0x0 0 > > > > > > r14 0x7fffffffd8d0 140737488345296 > > > > > > r15 0x7fffffffdcb8 140737488346296 > > > > > > rip 0x7ffff7ff5498 0x7ffff7ff5498 <decode_vec+21> > > > > > > eflags 0x10246 [ PF ZF IF RF ] > > > > > > cs 0x33 51 > > > > > > ss 0x2b 43 > > > > > > ds 0x0 0 > > > > > > es 0x0 0 > > > > > > fs 0x0 0 > > > > > > gs 0x0 0 > > > > > > > > > Disassembly of decode_dyn: > > > > > > > > > 0x00007ffff7ff58af <+0>: push %r14 > > > > > > 0x00007ffff7ff58b1 <+2>: push %rbx > > > > > > 0x00007ffff7ff58b2 <+3>: sub $0x108,%rsp > > > > > > 0x00007ffff7ff58b9 <+10>: mov %rdi,%rbx > > > > > > 0x00007ffff7ff58bc <+13>: mov 0x10(%rdi),%rdi > > > > > > 0x00007ffff7ff58c0 <+17>: mov %rsp,%r14 > > > > > > 0x00007ffff7ff58c3 <+20>: mov %r14,%rsi > > > > > > 0x00007ffff7ff58c6 <+23>: call 0x7ffff7ff5483 <decode_vec> > > > > > > 0x00007ffff7ff58cb <+28>: mov (%rbx),%rax > > > > > > > > > Disassembly of decode_vec: > > > > > > > > > 0x00007ffff7ff5483 <+0>: xor %eax,%eax > > > > > > 0x00007ffff7ff5485 <+2>: cmp $0x20,%rax > > > > > > 0x00007ffff7ff5489 <+6>: je 0x7ffff7ff5495 <decode_vec+18> > > > > > > 0x00007ffff7ff548b <+8>: andq $0x0,(%rsi,%rax,8) > > > > > > 0x00007ffff7ff5490 <+13>: inc %rax > > > > > > 0x00007ffff7ff5493 <+16>: jmp 0x7ffff7ff5485 <decode_vec+2> > > > > > > 0x00007ffff7ff5495 <+18>: push $0x1 > > > > > > 0x00007ffff7ff5497 <+20>: pop %rax > > > > > > => 0x00007ffff7ff5498 <+21>: mov (%rdi),%rcx > > > > > > 0x00007ffff7ff549b <+24>: test %rcx,%rcx > > > > > > 0x00007ffff7ff549e <+27>: je 0x7ffff7ff54c3 <decode_vec+64> > > > > > > 0x00007ffff7ff54a0 <+29>: lea -0x1(%rcx),%rdx > > > > > > 0x00007ffff7ff54a4 <+33>: cmp $0x1e,%rdx > > > > > > 0x00007ffff7ff54a8 <+37>: ja 0x7ffff7ff54bd <decode_vec+58> > > > > > > 0x00007ffff7ff54aa <+39>: shlx %rcx,%rax,%rcx > > > > > > 0x00007ffff7ff54af <+44>: or %rcx,(%rsi) > > > > > > 0x00007ffff7ff54b2 <+47>: mov (%rdi),%rcx > > > > > > 0x00007ffff7ff54b5 <+50>: mov 0x8(%rdi),%rdx > > > > > > 0x00007ffff7ff54b9 <+54>: mov %rdx,(%rsi,%rcx,8) > > > > > > 0x00007ffff7ff54bd <+58>: add $0x10,%rdi > > > > > > 0x00007ffff7ff54c1 <+62>: jmp 0x7ffff7ff5498 <decode_vec+21> > > > > > > 0x00007ffff7ff54c3 <+64>: ret > > > > > > > > > Jiahao XU > > > > > > Get Outlook for iOS<https://aka.ms/o0ukef> > > > ________________________________ > > > From: Rich Felker <dalias@libc.org> > > > Sent: Sunday, January 31, 2021 10:30:12 AM > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > On Sat, Jan 30, 2021 at 11:04:32PM +0000, Jiahao XU wrote: > > > > > So something like (in config.mak): > > > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > > > Thanks, with this I was able to build libc.so successfully with clang and created a 3.5 KB hello world program using clang and lld. > > > > > > > > However, I still wasn’t able to statically linked with libc. > > > > > > > > Once I added ‘-static’ to the compiler flags, the executable failed with ‘Segmentation fault (core dumped)’. > > > > > > It's libc.a, not libc.so, that will be involved in making a > > > static-linked binary. It's hard to know what's going wrong without > > > more information. Can you run under a debugger and provide a > > > backtrace, disassembly, and register dump for where the crash occurs? > > > > > > Rich > > > > > > > > > > ________________________________ > > > > From: Rich Felker <dalias@libc.org> > > > > Sent: Sunday, January 31, 2021 7:12:31 AM > > > > To: Jiahao XU <Jiahao_XU@outlook.com> > > > > Cc: musl@lists.openwall.com <musl@lists.openwall.com> > > > > Subject: Re: [musl] Can’t build musl with lto=thin > > > > > > > > On Fri, Jan 29, 2021 at 12:19:42PM +0000, Jiahao XU wrote: > > > > > musl-1.2.2 compilation with clang-11 failed to build libc.so at the final linking stage: > > > > > > > > > > ld.lld: error: undefined hidden symbol: __dls2 > > > > > >>> referenced by ld-temp.o > > > > > >>> lto.tmp:(_dlstart_c) > > > > > >>> did you mean: __dls3 > > > > > >>> defined in: lto.tmp > > > > > > > > > > I am using CFLAGS=‘-march=native -mtune=native -Oz -flto > > > > > -fmerge-all-constants -fomit-frame-pointer’ and LDFLAGS=‘-flto > > > > ^^^^^^^^^^^^^^^^^^^^^ > > > > > > > > The -fmerge-all-constants option gives non-conforming language > > > > semantics and should not be used, but that's a separate issue. > > > > > > > > > -fuse-ld=lld -Wl,—plugin-opt=O3,-O3,—icf=safe’. > > > > > > > > > No configure option is supplied. > > > > > > > > Otherwise, it's a known issue that LTO misses references from asm > > > > (both top-level and in functions). I think dlstart.lo and a few other > > > > files should just be built with LTO disabled; any LTO-type > > > > optimization in code that runs at this stage is inherently invalid, > > > > anyway. So something like (in config.mak): > > > > > > > > obj/ldso/dlstart.lo: CFLAGS_ALL += -fno-lto > > > > > > > > Rich ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 17:53 ` Rich Felker @ 2021-02-01 18:06 ` Fangrui Song 2021-02-01 21:20 ` Jiahao XU 0 siblings, 1 reply; 21+ messages in thread From: Fangrui Song @ 2021-02-01 18:06 UTC (permalink / raw) To: Jiahao XU; +Cc: musl On Mon, Feb 1, 2021 at 9:53 AM Rich Felker <dalias@libc.org> wrote: > > On Mon, Feb 01, 2021 at 12:31:47AM +0000, Jiahao XU wrote: > > Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. > > > > After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. > > ..data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. > > > > > > $ bloaty gc-section-a.out -- no-gc-section.a.out > > > > FILE SIZE VM SIZE > > > > -------------- -------------- > > > > +18% +56 -3.2% -8 .data > > > > [NEW] +6 [NEW] +6 [LOAD #2 [RX]] > > > > [DEL] -4 -66.7% -8 [LOAD #4 [RW]] > > > > -72.7% -8 [ = ] 0 [Unmapped] > > > > -32.0% -64 [ = ] 0 .comment > > > > -99.4% -19.4Ki -99.7% -19.4Ki .rodata > > > > -88.0% -22.3Ki -88.2% -22.3Ki .text > > > > -89.4% -41.8Ki -88.5% -41.8Ki TOTAL > > What is included in your hello world? Mine, static linked normally (no > LTO) is around 4k of text and virtually no rodata. (This is with GCC; > I'm not using clang.) If I compile with -fno-builtin so printf doesn't > get transformed to puts, there's about 16k of text and 3k of rodata, > but still nowhere near the ~20k you saw. > > Rich In LLD and LLVMgold.so's LTO configuration, -ffunction-sections & -fdata-sections are automatically enabled. Without them --gc-sections is not useful. -ffunction-sections & -fdata-sections are code generation options and not encoded in bitcode files. ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 18:06 ` Fangrui Song @ 2021-02-01 21:20 ` Jiahao XU 2021-02-13 17:54 ` Tim Tassonis 0 siblings, 1 reply; 21+ messages in thread From: Jiahao XU @ 2021-02-01 21:20 UTC (permalink / raw) To: Fangrui Song; +Cc: musl [-- Attachment #1: Type: text/plain, Size: 2249 bytes --] My hello.c: #include <stdio.h> int main() { puts(“Hello, world!”); return 0; } And if I changed puts to printf and uses -fno-builtin, it also generates a 16KB executable. Jiahao XU ________________________________ From: Fangrui Song <emacsray@gmail.com> Sent: Tuesday, February 2, 2021 5:06:26 AM To: Jiahao XU <Jiahao_XU@outlook.com> Cc: musl@lists.openwall.com <musl@lists.openwall.com> Subject: Re: [musl] Can’t build musl with lto=thin On Mon, Feb 1, 2021 at 9:53 AM Rich Felker <dalias@libc.org> wrote: > > On Mon, Feb 01, 2021 at 12:31:47AM +0000, Jiahao XU wrote: > > Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. > > > > After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. > > ..data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. > > > > > > $ bloaty gc-section-a.out -- no-gc-section.a.out > > > > FILE SIZE VM SIZE > > > > -------------- -------------- > > > > +18% +56 -3.2% -8 .data > > > > [NEW] +6 [NEW] +6 [LOAD #2 [RX]] > > > > [DEL] -4 -66.7% -8 [LOAD #4 [RW]] > > > > -72.7% -8 [ = ] 0 [Unmapped] > > > > -32.0% -64 [ = ] 0 .comment > > > > -99.4% -19.4Ki -99.7% -19.4Ki .rodata > > > > -88.0% -22.3Ki -88.2% -22.3Ki .text > > > > -89.4% -41.8Ki -88.5% -41.8Ki TOTAL > > What is included in your hello world? Mine, static linked normally (no > LTO) is around 4k of text and virtually no rodata. (This is with GCC; > I'm not using clang.) If I compile with -fno-builtin so printf doesn't > get transformed to puts, there's about 16k of text and 3k of rodata, > but still nowhere near the ~20k you saw. > > Rich In LLD and LLVMgold.so's LTO configuration, -ffunction-sections & -fdata-sections are automatically enabled. Without them --gc-sections is not useful. -ffunction-sections & -fdata-sections are code generation options and not encoded in bitcode files. [-- Attachment #2: Type: text/html, Size: 4339 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-01 21:20 ` Jiahao XU @ 2021-02-13 17:54 ` Tim Tassonis 2021-02-14 5:59 ` Markus Wichmann 0 siblings, 1 reply; 21+ messages in thread From: Tim Tassonis @ 2021-02-13 17:54 UTC (permalink / raw) To: musl On 2/1/21 10:20 PM, Jiahao XU wrote: > My hello.c: > #include <stdio.h> > int main() > { > puts(“Hello, world!”); > return 0; > } > > And if I changed puts to printf and uses -fno-builtin, it also generates > a 16KB executable. Strange, I get 19k without -fno-builtin and 28K with it: timtas@lfsda0:~$ cat hello.c #include <stdio.h> int main(int arc, char **argv) { printf("Hello World!\n"); return 0; } timtas@lfsda0:~$ musl-gcc -c hello.c -o hello.o timtas@lfsda0:~$ musl-gcc -o hello hello.o timtas@lfsda0:~$ ls -lh hello -rwxr-xr-x 1 timtas timtas 19K Feb 13 18:49 hello timtas@lfsda0:~$ musl-gcc -fno-builtin -c hello.c -o hello.o timtas@lfsda0:~$ musl-gcc -fno-builtin -o hello hello.o timtas@lfsda0:~$ ls -lh hello -rwxr-xr-x 1 timtas timtas 28K Feb 13 18:50 hello This is musl 1.2.1 with gcc 10.2.0 Bye Tim > > Jiahao XU > ------------------------------------------------------------------------ > *From:* Fangrui Song <emacsray@gmail.com> > *Sent:* Tuesday, February 2, 2021 5:06:26 AM > *To:* Jiahao XU <Jiahao_XU@outlook.com> > *Cc:* musl@lists.openwall.com <musl@lists.openwall.com> > *Subject:* Re: [musl] Can’t build musl with lto=thin > On Mon, Feb 1, 2021 at 9:53 AM Rich Felker <dalias@libc.org> wrote: >> >> On Mon, Feb 01, 2021 at 12:31:47AM +0000, Jiahao XU wrote: >> > Interesting enough, I found —gc-section used along with -flto can cut the size of final hello_world executable from 48KB to 5KB. >> > >> > After investigating with bloaty, I found that —gc-section along with -flto is able to cut .text from 25.4 KiB to 3.04 KiB, and cut the .rodata from 19.5 KiB to 120 bytes. >> > ..data section however, seen an increase from 316 bytes to 372 bytes, but the VM size is cut from 252 to 244 bytes. >> > >> > >> > $ bloaty gc-section-a.out -- no-gc-section.a.out >> > >> > FILE SIZE VM SIZE >> > >> > -------------- -------------- >> > >> > +18% +56 -3.2% -8 .data >> > >> > [NEW] +6 [NEW] +6 [LOAD #2 [RX]] >> > >> > [DEL] -4 -66.7% -8 [LOAD #4 [RW]] >> > >> > -72.7% -8 [ = ] 0 [Unmapped] >> > >> > -32.0% -64 [ = ] 0 .comment >> > >> > -99.4% -19.4Ki -99.7% -19.4Ki .rodata >> > >> > -88.0% -22.3Ki -88.2% -22.3Ki .text >> > >> > -89.4% -41.8Ki -88.5% -41.8Ki TOTAL >> >> What is included in your hello world? Mine, static linked normally (no >> LTO) is around 4k of text and virtually no rodata. (This is with GCC; >> I'm not using clang.) If I compile with -fno-builtin so printf doesn't >> get transformed to puts, there's about 16k of text and 3k of rodata, >> but still nowhere near the ~20k you saw. >> >> Rich > > In LLD and LLVMgold.so's LTO configuration, -ffunction-sections & > -fdata-sections are automatically enabled. > Without them --gc-sections is not useful. > > -ffunction-sections & -fdata-sections are code generation options and > not encoded in bitcode files. -- decentral.ch - IT Stuff Tim Tassonis Hohlstrasse 400 c/o Baubüro Insitu 8048 Zürich stuff@decentral.ch +41 79 229 36 17 ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [musl] Can’t build musl with lto=thin 2021-02-13 17:54 ` Tim Tassonis @ 2021-02-14 5:59 ` Markus Wichmann 0 siblings, 0 replies; 21+ messages in thread From: Markus Wichmann @ 2021-02-14 5:59 UTC (permalink / raw) To: musl On Sat, Feb 13, 2021 at 06:54:24PM +0100, Tim Tassonis wrote: > On 2/1/21 10:20 PM, Jiahao XU wrote: > > And if I changed puts to printf and uses -fno-builtin, it also generates > > a 16KB executable. > > Strange, I get 19k without -fno-builtin and 28K with it: > GCC detects the use of printf() with only a constant string ending in \n and replaces it with a call to puts(). You can see that by looking at the disassembly, or making GCC generate the assembly. Ciao, Markus ^ permalink raw reply [flat|nested] 21+ messages in thread
end of thread, other threads:[~2021-02-14 5:59 UTC | newest] Thread overview: 21+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2021-01-29 12:19 [musl] Can’t build musl with lto=thin Jiahao XU 2021-01-29 20:04 ` Szabolcs Nagy 2021-01-29 20:53 ` Fangrui Song 2021-01-30 20:12 ` Rich Felker 2021-01-30 20:59 ` Леонид Юрьев (Leonid Yuriev) [not found] ` <SYBP282MB020277E08EB49C9AC9BF90888AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM> 2021-01-30 23:30 ` Rich Felker [not found] ` <SYBP282MB0202BF16DD08A2F1537FD5428AB89@SYBP282MB0202.AUSP282.PROD.OUTLOOK.COM> 2021-01-31 5:01 ` Rich Felker 2021-01-31 5:32 ` Jiahao XU 2021-01-31 21:01 ` Rich Felker 2021-01-31 22:29 ` Jiahao XU 2021-02-01 0:22 ` Jiahao XU 2021-02-01 0:31 ` Jiahao XU 2021-02-01 0:44 ` Jiahao XU 2021-02-01 0:50 ` Jiahao XU 2021-02-01 13:20 ` Jiahao XU 2021-02-01 16:55 ` Nick Bray 2021-02-01 17:53 ` Rich Felker 2021-02-01 18:06 ` Fangrui Song 2021-02-01 21:20 ` Jiahao XU 2021-02-13 17:54 ` Tim Tassonis 2021-02-14 5:59 ` Markus Wichmann
Code repositories for project(s) associated with this public inbox https://git.vuxu.org/mirror/musl/ This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).