Starting with Rust 1.87.0 (expected 2025-05-15), `objtool` may report:
rust/core.o: warning: objtool: _R..._4core9panicking9panic_fmt() falls
through to next function _R..._4core9panicking18panic_nounwind_fmt()
rust/core.o: warning: objtool: _R..._4core9panicking18panic_nounwind_fmt()
falls through to next function _R..._4core9panicking5panic()
The reason is that `rust_begin_unwind` is now mangled:
_R..._7___rustc17rust_begin_unwind
Thus add the mangled one to the list so that `objtool` knows it is
actually `noreturn`.
See commit 56d680dd23c3 ("objtool/rust: list `noreturn` Rust functions")
for more details.
Alternatively, we could remove the fixed one in `noreturn.h` and relax
this test to cover both, but it seems best to be strict as long as we can.
Cc: stable@vger.kernel.org # Needed in 6.12.y and later (Rust is pinned in older LTSs).
Cc: Josh Poimboeuf <jpoimboe@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
---
tools/objtool/check.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/tools/objtool/check.c b/tools/objtool/check.c
index 3a411064fa34..b21b12ec88d9 100644
--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -227,6 +227,7 @@ static bool is_rust_noreturn(const struct symbol *func)
str_ends_with(func->name, "_4core9panicking19assert_failed_inner") ||
str_ends_with(func->name, "_4core9panicking30panic_null_pointer_dereference") ||
str_ends_with(func->name, "_4core9panicking36panic_misaligned_pointer_dereference") ||
+ str_ends_with(func->name, "_7___rustc17rust_begin_unwind") ||
strstr(func->name, "_4core9panicking13assert_failed") ||
strstr(func->name, "_4core9panicking11panic_const24panic_const_") ||
(strstr(func->name, "_4core5slice5index24slice_") &&
--
2.49.0Hello Miguel,
On Fri, May 02, 2025 at 04:02:33PM +0200, Miguel Ojeda wrote:
> Starting with Rust 1.87.0 (expected 2025-05-15), `objtool` may report:
>
> rust/core.o: warning: objtool: _R..._4core9panicking9panic_fmt() falls
> through to next function _R..._4core9panicking18panic_nounwind_fmt()
>
> rust/core.o: warning: objtool: _R..._4core9panicking18panic_nounwind_fmt()
> falls through to next function _R..._4core9panicking5panic()
We are seeing a similar issue with the patch [1]:
RUSTC [M] drivers/gpu/nova-core/nova_core.o
drivers/gpu/nova-core/nova_core.o: warning: objtool:
<nova_core::vbios::PciAtBiosImage as
core::convert::TryFrom<nova_core::vbios::BiosImageBase>>::try_from() falls
through to next function <nova_core::vbios::FwSecBiosImage>::fwsec_header()
The code in concern is implementing try_from():
+
+impl TryFrom<BiosImageBase> for PciAtBiosImage {
+ type Error = Error;
+
+ fn try_from(base: BiosImageBase) -> Result<Self> {
I dumped the codegen [2] for this function and at the end of the codegen, there
is a call instruction to to the fwsec_header() function.
Any thoughts on how to fix the warning?
thanks,
- Joel
[1] https://lore.kernel.org/all/20250420-nova-frts-v1-13-ecd1cca23963@nvidia.com/
[2] https://paste.debian.net/1374516/
> The reason is that `rust_begin_unwind` is now mangled:
>
> _R..._7___rustc17rust_begin_unwind
>
> Thus add the mangled one to the list so that `objtool` knows it is
> actually `noreturn`.
>
> See commit 56d680dd23c3 ("objtool/rust: list `noreturn` Rust functions")
> for more details.
>
> Alternatively, we could remove the fixed one in `noreturn.h` and relax
> this test to cover both, but it seems best to be strict as long as we can.
>
> Cc: stable@vger.kernel.org # Needed in 6.12.y and later (Rust is pinned in older LTSs).
> Cc: Josh Poimboeuf <jpoimboe@kernel.org>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
> ---
> tools/objtool/check.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/tools/objtool/check.c b/tools/objtool/check.c
> index 3a411064fa34..b21b12ec88d9 100644
> --- a/tools/objtool/check.c
> +++ b/tools/objtool/check.c
> @@ -227,6 +227,7 @@ static bool is_rust_noreturn(const struct symbol *func)
> str_ends_with(func->name, "_4core9panicking19assert_failed_inner") ||
> str_ends_with(func->name, "_4core9panicking30panic_null_pointer_dereference") ||
> str_ends_with(func->name, "_4core9panicking36panic_misaligned_pointer_dereference") ||
> + str_ends_with(func->name, "_7___rustc17rust_begin_unwind") ||
> strstr(func->name, "_4core9panicking13assert_failed") ||
> strstr(func->name, "_4core9panicking11panic_const24panic_const_") ||
> (strstr(func->name, "_4core5slice5index24slice_") &&
> --
> 2.49.0
On Tue, May 13, 2025 at 8:08 PM Joel Fernandes <joelagnelf@nvidia.com> wrote: > > Any thoughts on how to fix the warning? For future reference, this gets fixed with https://lore.kernel.org/rust-for-linux/20250520185555.825242-1-ojeda@kernel.org/ I hope that helps (and apologies for the delay, a bunch of us were in RustWeek :) Cheers, Miguel
On Tue, May 13, 2025 at 02:07:57PM -0400, Joel Fernandes wrote:
> Hello Miguel,
>
> On Fri, May 02, 2025 at 04:02:33PM +0200, Miguel Ojeda wrote:
> > Starting with Rust 1.87.0 (expected 2025-05-15), `objtool` may report:
> >
> > rust/core.o: warning: objtool: _R..._4core9panicking9panic_fmt() falls
> > through to next function _R..._4core9panicking18panic_nounwind_fmt()
> >
> > rust/core.o: warning: objtool: _R..._4core9panicking18panic_nounwind_fmt()
> > falls through to next function _R..._4core9panicking5panic()
>
> We are seeing a similar issue with the patch [1]:
>
> RUSTC [M] drivers/gpu/nova-core/nova_core.o
> drivers/gpu/nova-core/nova_core.o: warning: objtool:
> <nova_core::vbios::PciAtBiosImage as
> core::convert::TryFrom<nova_core::vbios::BiosImageBase>>::try_from() falls
> through to next function <nova_core::vbios::FwSecBiosImage>::fwsec_header()
>
> The code in concern is implementing try_from():
> +
> +impl TryFrom<BiosImageBase> for PciAtBiosImage {
> + type Error = Error;
> +
> + fn try_from(base: BiosImageBase) -> Result<Self> {
>
> I dumped the codegen [2] for this function and at the end of the codegen, there
> is a call instruction to to the fwsec_header() function.
>
> Any thoughts on how to fix the warning?
Btw, Danilo mentioned to me the latest Rust compiler (1.86?) does not give
this warning for that patch.
Mine is on 1.85. So if anyone else other than me is suffering from this
warning, do upgrade. :-)
thanks,
- Joel
On 5/13/25 2:58 PM, Joel Fernandes wrote: > On Tue, May 13, 2025 at 02:07:57PM -0400, Joel Fernandes wrote: >> On Fri, May 02, 2025 at 04:02:33PM +0200, Miguel Ojeda wrote: >>> Starting with Rust 1.87.0 (expected 2025-05-15), `objtool` may report: ... > Btw, Danilo mentioned to me the latest Rust compiler (1.86?) does not give > this warning for that patch. I'm sorry to burst this happy bubble, but I just upgraded to rustc 1.86 and did a clean build, and I *am* setting these warnings: $ rustc --version rustc 1.86.0 (05f9846f8 2025-03-31) (Arch Linux rust 1:1.86.0-1) ... RUSTC L rust/kernel.o rust/kernel.o: warning: objtool: _RNvMNtCsbA27Kl4nY2_6kernel6deviceNtB2_6Device10get_device() falls through to next function _RNvXs4_NtCsbA27Kl4nY2_6kernel9auxiliaryNtB5_6DeviceNtNtB7_5types16AlwaysRefCounted7inc_ref() rust/kernel.o: warning: objtool: _RNvMs2_NtNtCsbA27Kl4nY2_6kernel2fs4fileNtB5_9LocalFile4cred() falls through to next function _RNvXs8_NtNtCsbA27Kl4nY2_6kernel2fs4fileNtB5_10BadFdErrorNtNtCs8y00iZOEpTQ_4core3fmt5Debug3fmt() ...followed by 10 or 12 more of the same "falls through" type. > > Mine is on 1.85. So if anyone else other than me is suffering from this > warning, do upgrade. :-) > Looks like that might not suffice! thanks, -- John Hubbard
On Tue, 2025-05-13 at 17:22 -0700, John Hubbard wrote: > On 5/13/25 2:58 PM, Joel Fernandes wrote: > > On Tue, May 13, 2025 at 02:07:57PM -0400, Joel Fernandes wrote: > > > On Fri, May 02, 2025 at 04:02:33PM +0200, Miguel Ojeda wrote: > > > > Starting with Rust 1.87.0 (expected 2025-05-15), `objtool` may report: > ... > > Btw, Danilo mentioned to me the latest Rust compiler (1.86?) does not give > > this warning for that patch. > > I'm sorry to burst this happy bubble, but I just upgraded to rustc 1.86 and did > a clean build, and I *am* setting these warnings: I see these warnings with .c code also: CHK kernel/kheaders_data.tar.xz drivers/media/pci/solo6x10/solo6x10-tw28.o: error: objtool: tw28_set_ctrl_val() falls through to next function tw28_get_ctrl_val() make[9]: *** [scripts/Makefile.build:203: drivers/media/pci/solo6x10/solo6x10-tw28.o] Error 1 I think it's an objtool bug and not a rustc bug.
On 5/13/2025 8:43 PM, Timur Tabi wrote: > On Tue, 2025-05-13 at 17:22 -0700, John Hubbard wrote: >> On 5/13/25 2:58 PM, Joel Fernandes wrote: >>> On Tue, May 13, 2025 at 02:07:57PM -0400, Joel Fernandes wrote: >>>> On Fri, May 02, 2025 at 04:02:33PM +0200, Miguel Ojeda wrote: >>>>> Starting with Rust 1.87.0 (expected 2025-05-15), `objtool` may report: >> ... >>> Btw, Danilo mentioned to me the latest Rust compiler (1.86?) does not give >>> this warning for that patch. >> >> I'm sorry to burst this happy bubble, but I just upgraded to rustc 1.86 and did >> a clean build, and I *am* setting these warnings: > > I see these warnings with .c code also: > > CHK kernel/kheaders_data.tar.xz > drivers/media/pci/solo6x10/solo6x10-tw28.o: error: objtool: tw28_set_ctrl_val() falls through to > next function tw28_get_ctrl_val() > make[9]: *** [scripts/Makefile.build:203: drivers/media/pci/solo6x10/solo6x10-tw28.o] Error 1 > > I think it's an objtool bug and not a rustc bug. Thanks John and Timur. And sigh, fwiw I pulled the latest rust nightly build and I see the warning as well: rustc --version rustc 1.89.0-nightly (414482f6a 2025-05-13) I am leaning more towards Timur's opinion that this is more than likely an objtool issue. - Joel
On Wed, May 14, 2025 at 10:52:17AM -0400, Joel Fernandes wrote: > On 5/13/2025 8:43 PM, Timur Tabi wrote: > > On Tue, 2025-05-13 at 17:22 -0700, John Hubbard wrote: > >> On 5/13/25 2:58 PM, Joel Fernandes wrote: > >>> On Tue, May 13, 2025 at 02:07:57PM -0400, Joel Fernandes wrote: > >>>> On Fri, May 02, 2025 at 04:02:33PM +0200, Miguel Ojeda wrote: > >>>>> Starting with Rust 1.87.0 (expected 2025-05-15), `objtool` may report: > >> ... > >>> Btw, Danilo mentioned to me the latest Rust compiler (1.86?) does not give > >>> this warning for that patch. > >> > >> I'm sorry to burst this happy bubble, but I just upgraded to rustc 1.86 and did > >> a clean build, and I *am* setting these warnings: > > > > I see these warnings with .c code also: > > > > CHK kernel/kheaders_data.tar.xz > > drivers/media/pci/solo6x10/solo6x10-tw28.o: error: objtool: tw28_set_ctrl_val() falls through to > > next function tw28_get_ctrl_val() > > make[9]: *** [scripts/Makefile.build:203: drivers/media/pci/solo6x10/solo6x10-tw28.o] Error 1 > > > > I think it's an objtool bug and not a rustc bug. > > Thanks John and Timur. > And sigh, fwiw I pulled the latest rust nightly build and I see the warning as well: > > rustc --version > rustc 1.89.0-nightly (414482f6a 2025-05-13) > > I am leaning more towards Timur's opinion that this is more than likely an > objtool issue. The above warning is in completely different code from the Rust one, so they're likely unrelated. The fallthrough warnings are typically caused by either Clang undefined behavior (usually potential divide by zero or negative shift), or a call to an unannotated noreturn function. Timur, can you share your .config and compiler version? -- Josh
On Wed, 2025-05-14 at 12:14 -0700, Josh Poimboeuf wrote: > The above warning is in completely different code from the Rust one, so > they're likely unrelated. True, but the fall-through is bogus in the C code as well. > The fallthrough warnings are typically caused by either Clang undefined > behavior (usually potential divide by zero or negative shift), or a call > to an unannotated noreturn function. > > Timur, can you share your .config and compiler version? .config: https://pastebin.com/inDHfmbG ttabi@ttabi:~$ llvm-config --version 18.1.3 ttabi@ttabi:~$ gcc --version gcc (Ubuntu 14.2.0-4ubuntu2~24.04) 14.2.0 Since I build with LLVM=1, I'm assuming the answer is 18.1.3
On Wed, May 14, 2025 at 07:46:49PM +0000, Timur Tabi wrote: > On Wed, 2025-05-14 at 12:14 -0700, Josh Poimboeuf wrote: > > The above warning is in completely different code from the Rust one, so > > they're likely unrelated. > > True, but the fall-through is bogus in the C code as well. > > > The fallthrough warnings are typically caused by either Clang undefined > > behavior (usually potential divide by zero or negative shift), or a call > > to an unannotated noreturn function. > > > > Timur, can you share your .config and compiler version? > > .config: https://pastebin.com/inDHfmbG > > ttabi@ttabi:~$ llvm-config --version > 18.1.3 > ttabi@ttabi:~$ gcc --version > gcc (Ubuntu 14.2.0-4ubuntu2~24.04) 14.2.0 > > Since I build with LLVM=1, I'm assuming the answer is 18.1.3 I'm not able to recreate, can you run with OBJTOOL_VERBOSE=1 and paste the output? -- Josh
On Thu, 2025-05-15 at 09:18 -0700, Josh Poimboeuf wrote: > > Since I build with LLVM=1, I'm assuming the answer is 18.1.3 > > I'm not able to recreate, can you run with OBJTOOL_VERBOSE=1 and paste > the output? You probably can't repro because it includes code that hasn't been merged upstream yet. Try this: https://github.com/ttabi/linux/commits/alex CHK kernel/kheaders_data.tar.xz drivers/gpu/nova-core/nova_core.o: warning: objtool: _RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFro mNtB5_13BiosImageBaseE8try_from() falls through to next function _RNvMsd_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14FwSecBiosImage12fwsec_header() drivers/gpu/nova-core/nova_core.o: warning: objtool: _RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFro mNtB5_13BiosImageBaseE8try_from+0x5c: (branch) drivers/gpu/nova-core/nova_core.o: warning: objtool: _RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFro mNtB5_13BiosImageBaseE8try_from+0x4a: (branch) drivers/gpu/nova-core/nova_core.o: warning: objtool: _RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFro mNtB5_13BiosImageBaseE8try_from+0x0: <=== (sym) ./tools/objtool/objtool --hacks=jump_label --hacks=noinstr --hacks=skylake --mcount --mnop -- retpoline --rethunk --sls --stackval --static-call --uaccess --prefix=16 --module drivers/gpu/nova- core/nova_core.o.orig -o drivers/gpu/nova-core/nova_core.o 0000 0000000000006230 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from>: 0000 6230: 55 push %rbp 0001 6231: 48 89 e5 mov %rsp,%rbp 0004 6234: 53 push %rbx 0005 6235: 48 83 ec 18 sub $0x18,%rsp 0009 6239: 48 89 fb mov %rdi,%rbx 000c 623c: 48 8b 3e mov (%rsi),%rdi 000f 623f: 48 8b 4e 10 mov 0x10(%rsi),%rcx 0013 6243: c7 45 f2 ff b8 42 49 movl $0x4942b8ff,-0xe(%rbp) 001a 624a: 66 c7 45 f6 54 00 movw $0x54,-0xa(%rbp) 0020 6250: 48 83 f9 06 cmp $0x6,%rcx 0024 6254: 72 42 jb 6298 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x68> 0026 6256: 48 8d 51 fb lea -0x5(%rcx),%rdx 002a 625a: 31 c0 xor %eax,%eax 002c 625c: 0f 1f 40 00 nopl 0x0(%rax) 0030 6260: 44 8b 04 07 mov (%rdi,%rax,1),%r8d 0034 6264: 44 33 45 f2 xor -0xe(%rbp),%r8d 0038 6268: 44 0f b7 4c 07 04 movzwl 0x4(%rdi,%rax,1),%r9d 003e 626e: 66 44 33 4d f6 xor -0xa(%rbp),%r9w 0043 6273: 45 0f b7 c9 movzwl %r9w,%r9d 0047 6277: 45 09 c1 or %r8d,%r9d 004a 627a: 74 0a je 6286 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x56> 004c 627c: 48 ff c0 inc %rax 004f 627f: 48 39 c2 cmp %rax,%rdx 0052 6282: 75 dc jne 6260 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x30> 0054 6284: eb 12 jmp 6298 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x68> 0056 6286: 48 89 ca mov %rcx,%rdx 0059 6289: 48 29 c2 sub %rax,%rdx 005c 628c: 0f 82 ac 00 00 00 jb 633e <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x10e> 0062 6292: 48 83 fa 0c cmp $0xc,%rdx 0066 6296: 73 30 jae 62c8 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x98> 0068 6298: c7 43 08 ea ff ff ff movl $0xffffffea,0x8(%rbx) 006f 629f: 48 c7 03 02 00 00 00 movq $0x2,(%rbx) 0076 62a6: 48 8b 46 08 mov 0x8(%rsi),%rax 007a 62aa: 48 85 c0 test %rax,%rax 007d 62ad: 48 0f 44 f8 cmove %rax,%rdi 0081 62b1: 31 f6 xor %esi,%esi 0083 62b3: 31 d2 xor %edx,%edx 0085 62b5: e8 00 00 00 00 call 62ba <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x8a> 62b6: R_X86_64_PLT32 rust_helper_krealloc-0x4 008a 62ba: 48 89 d8 mov %rbx,%rax 008d 62bd: 48 83 c4 18 add $0x18,%rsp 0091 62c1: 5b pop %rbx 0092 62c2: 5d pop %rbp 0093 62c3: e9 00 00 00 00 jmp 62c8 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x98> 62c4: R_X86_64_PLT32 __x86_return_thunk-0x4 0098 62c8: 66 81 3c 07 ff b8 cmpw $0xb8ff,(%rdi,%rax,1) 009e 62ce: 75 c8 jne 6298 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x68> 00a0 62d0: 81 7c 07 02 42 49 54 00 cmpl $0x544942,0x2(%rdi,%rax,1) 00a8 62d8: 75 be jne 6298 <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x68> 00aa 62da: 0f b6 4c 07 0a movzbl 0xa(%rdi,%rax,1),%ecx 00af 62df: 8b 54 07 06 mov 0x6(%rdi,%rax,1),%edx 00b3 62e3: 44 0f b6 44 07 0b movzbl 0xb(%rdi,%rax,1),%r8d 00b9 62e9: c7 45 ec 42 49 54 00 movl $0x544942,-0x14(%rbp) 00c0 62f0: 66 c7 45 f0 ff b8 movw $0xb8ff,-0x10(%rbp) 00c6 62f6: 48 c1 e1 20 shl $0x20,%rcx 00ca 62fa: 49 c1 e0 28 shl $0x28,%r8 00ce 62fe: 49 09 c8 or %rcx,%r8 00d1 6301: 44 8b 4d ee mov -0x12(%rbp),%r9d 00d5 6305: 48 c7 03 01 00 00 00 movq $0x1,(%rbx) 00dc 630c: 48 89 43 08 mov %rax,0x8(%rbx) 00e0 6310: 48 8d 7b 10 lea 0x10(%rbx),%rdi 00e4 6314: b9 0d 00 00 00 mov $0xd,%ecx 00e9 6319: f3 48 a5 rep movsq %ds:(%rsi),%es:(%rdi) 00ec 631c: c7 43 78 01 00 42 49 movl $0x49420001,0x78(%rbx) 00f3 6323: 44 89 4b 7c mov %r9d,0x7c(%rbx) 00f7 6327: 49 c1 e8 20 shr $0x20,%r8 00fb 632b: 66 44 89 83 84 00 00 00 mov %r8w,0x84(%rbx) 0103 6333: 89 93 80 00 00 00 mov %edx,0x80(%rbx) 0109 6339: e9 7c ff ff ff jmp 62ba <_RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFr omNtB5_13BiosImageBaseE8try_from+0x8a> 010e 633e: 48 89 c7 mov %rax,%rdi 0111 6341: 48 89 ce mov %rcx,%rsi 0114 6344: 48 c7 c2 00 00 00 00 mov $0x0,%rdx 6347: R_X86_64_32S .rodata+0x1058 011b 634b: e8 00 00 00 00 call 6350 <.Ltmp38> 634c: R_X86_64_PLT32 _RNvNtNtCsgK88DPai1lC_4core5slice5index26slice_start_index_len_fail-0x4
On Thu, May 15, 2025 at 07:06:10PM +0000, Timur Tabi wrote: > On Thu, 2025-05-15 at 09:18 -0700, Josh Poimboeuf wrote: > > > Since I build with LLVM=1, I'm assuming the answer is 18.1.3 > > > > I'm not able to recreate, can you run with OBJTOOL_VERBOSE=1 and paste > > the output? > > You probably can't repro because it includes code that hasn't been merged upstream yet. Try this: > > https://github.com/ttabi/linux/commits/alex > > CHK kernel/kheaders_data.tar.xz > drivers/gpu/nova-core/nova_core.o: warning: objtool: > _RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFro I'm not yet qualified to look at the rust warning, I was actually wondering about the C one you reported: drivers/media/pci/solo6x10/solo6x10-tw28.o: error: objtool: tw28_set_ctrl_val() falls through to next function tw28_get_ctrl_val() -- Josh
On Thu, 2025-05-15 at 15:16 -0700, Josh Poimboeuf wrote: > I'm not yet qualified to look at the rust warning, I was actually > wondering about the C one you reported: > > drivers/media/pci/solo6x10/solo6x10-tw28.o: error: objtool: tw28_set_ctrl_val() falls through to > next function tw28_get_ctrl_val() https://pastebin.com/DYuifnTB
On Thu, 15 May 2025 19:06:10 +0000, Timur Tabi <ttabi@nvidia.com> wrote:
> On Thu, 2025-05-15 at 09:18 -0700, Josh Poimboeuf wrote:
> > > Since I build with LLVM=1, I'm assuming the answer is 18.1.3
> >
> > I'm not able to recreate, can you run with OBJTOOL_VERBOSE=1 and paste
> > the output?
>
> You probably can't repro because it includes code that hasn't been merged upstream yet. Try this:
>
> https://github.com/ttabi/linux/commits/alex
>
> CHK kernel/kheaders_data.tar.xz
> drivers/gpu/nova-core/nova_core.o: warning: objtool:
> _RNvXsa_NtCs8S3917Wilyo_9nova_core5vbiosNtB5_14PciAtBiosImageINtNtCsgK88DPai1lC_4core7convert7TryFro
> mNtB5_13BiosImageBaseE8try_from() falls through to next function
...
> 011b 634b: e8 00 00 00 00 call 6350 <.Ltmp38> 634c:
> R_X86_64_PLT32
> _RNvNtNtCsgK88DPai1lC_4core5slice5index26slice_start_index_len_fail-0x4
Yup, that's an unrecognized noreturn function.
src/core/slice/index.rs:
> #[cfg_attr(not(feature = "panic_immediate_abort"), inline(never), cold)]
> #[cfg_attr(feature = "panic_immediate_abort", inline)]
> #[track_caller]
> #[rustc_const_unstable(feature = "const_slice_index", issue = "none")]
> const fn slice_start_index_len_fail(index: usize, len: usize) -> ! {
> // SAFETY: we are just panicking here
> unsafe {
> const_eval_select(
> (index, len),
> slice_start_index_len_fail_ct,
> slice_start_index_len_fail_rt,
> )
> }
> }
>
> // FIXME const-hack
> #[inline]
> #[track_caller]
> fn slice_start_index_len_fail_rt(index: usize, len: usize) -> ! {
> panic!("range start index {index} out of range for slice of length {len}");
> }
The return is at offset 0093 (x62c3) with a jump to __x86_return_thunk.
© 2016 - 2026 Red Hat, Inc.