https://sourceware.org/bugzilla/show_bug.cgi?id=32809
xiaoqiang-wang changed:
What|Removed |Added
CC||xiaoqiang-wang at outlook dot
com
-
Replying to this email means your email address will be shared with the
team that works on this product.
https://issues.oss-fuzz.com/issues/432531556
Reference Info: 432531556 binutils:fuzz_as: Undefined-shift in
read_a_source_file
component: Public Trackers > 1362134 > OSS Fuzz
status: New
rep
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #8 from Indu Bhagat ---
Created attachment 16188
--> https://sourceware.org/bugzilla/attachment.cgi?id=16188&action=edit
Possible resolution
At this time, I think its not unreasonable to assert for cie_fre (like the way
it is ri
https://sourceware.org/bugzilla/show_bug.cgi?id=33166
--- Comment #4 from Indu Bhagat ---
Right. Binutils 2.44 shows the assertion, but this should be fixed in Binutils
2.45.
--
You are receiving this mail because:
You are on the CC list for the bug.
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #7 from Indu Bhagat ---
The assertion fail is because the SFrame generation code assumes that a
DW_CFA_restore must be restoring the state of the register to the one at the
beginning of the function. But in this case, the input as
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #6 from Indu Bhagat ---
Created attachment 16187
--> https://sourceware.org/bugzilla/attachment.cgi?id=16187&action=edit
offending function as 33170_testcase.s
Confirmed.
The offending function is attached as 33170_testcase.s
https://sourceware.org/bugzilla/show_bug.cgi?id=33168
--- Comment #2 from Indu Bhagat ---
AC_USE_SYSTEM_EXTENSIONS is used in other configure.ac in Binutils too. So, I
am not sure if removing it from libsframe is ideal. I dont have a good grasp
of the matter here (cannot fully evaluate if this
https://sourceware.org/bugzilla/show_bug.cgi?id=33162
--- Comment #14 from Nick Alcock ---
I think I have something which can literally check to see whether the linker is
dedupping CTF or simply concatenating it :) so even non-GNU linkers should just
skip libctf tests that require a deduplicating
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #5 from Sam James ---
$ gas/as-new /tmp/SymbolCache.s --gsframe
/tmp/SymbolCache.s: Assembler messages:
/tmp/SymbolCache.s: Warning: no SFrame FDE emitted; .cfi_escape with op (0x2e)
/tmp/SymbolCache.s: Warning: no SFrame FDE emitt
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #4 from Sam James ---
```
(rr) bt
#0 sframe_xlate_do_restore (xlate_ctx=0x55af5ff417f0, cfi_insn=0x55af601f09f0)
at gen-sframe.c:1291
#1 0x55af54443b65 in sframe_do_cfi_insn (xlate_ctx=0x55af5ff417f0,
cfi_insn=0x55af601f09f0)
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #3 from Sam James ---
The line numbers are slightly wrong as I have warnings patched out, sorry. Let
me get real ones from master.
--
You are receiving this mail because:
You are on the CC list for the bug.
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
Bug ID: 33170
Summary: Internal error in sframe_xlate_do_restore when
building llvm-19.1.7
Product: binutils
Version: 2.46 (HEAD)
Status: NEW
Severity: norma
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #2 from Sam James ---
```
$ gdb --args as SymbolCache.s
[...]
(gdb) b as_abort
Breakpoint 1 at 0x55581570: file
/usr/src/debug/sys-devel/binutils-/binutils/gas/messages.c, line 347.
Breakpoint 1, as_abort (file=0x555db0
https://sourceware.org/bugzilla/show_bug.cgi?id=33170
--- Comment #1 from Sam James ---
Created attachment 16186
--> https://sourceware.org/bugzilla/attachment.cgi?id=16186&action=edit
SymbolCache.s.xz
--
You are receiving this mail because:
You are on the CC list for the bug.
https://sourceware.org/bugzilla/show_bug.cgi?id=33169
Bug ID: 33169
Summary: Some pr25617-1 tests FAIL on Solaris
Product: binutils
Version: 2.45
Status: NEW
Severity: normal
Priority: P2
Component: ld
https://sourceware.org/bugzilla/show_bug.cgi?id=33168
--- Comment #1 from Rainer Orth ---
> * A patch like the following:
>
> diff --git a/libsframe/configure.ac b/libsframe/configure.ac
> --- a/libsframe/configure.ac
> +++ b/libsframe/configure.ac
> @@ -23,6 +23,13 @@ AC_CONFIG_SRCDIR(sframe.c)
https://sourceware.org/bugzilla/show_bug.cgi?id=33162
--- Comment #13 from Rainer Orth ---
> --- Comment #4 from Nick Alcock ---
> We can probably assume that pre-v8plus hardware is long extinct, so this
> horrible hack will probably do. (A shame we can't arrange to assemble "the
> same
> way G
https://sourceware.org/bugzilla/show_bug.cgi?id=33168
Bug ID: 33168
Summary: libsframe tests don't build on Solaris
Product: binutils
Version: 2.45
Status: NEW
Severity: normal
Priority: P2
Component: libsframe
18 matches
Mail list logo
