[Bug binutils/25135] Cannot link because of ".dynsym local symbol at index"
https://sourceware.org/bugzilla/show_bug.cgi?id=25135 Benjamin changed: What|Removed |Added Status|RESOLVED|UNCONFIRMED Resolution|DUPLICATE |--- --- Comment #2 from Benjamin --- The commit that fixes bug 24857 if from July: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=0231a51ef7ff49336d3a2f0e4eec09cd17b23c95 Tags binutils-2_33 and binutils-2_33_1 both contain this commit. I tested with versions binutils 2.33 (from eoan) and binutils 2.33.1 (from devel). I still have the issue with both versions. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/25136] New: mips64: set abi as N64 when the ABI section triple is gnuabi64
https://sourceware.org/bugzilla/show_bug.cgi?id=25136 Bug ID: 25136 Summary: mips64: set abi as N64 when the ABI section triple is gnuabi64 Product: binutils Version: unspecified Status: NEW Severity: normal Priority: P2 Component: binutils Assignee: unassigned at sourceware dot org Reporter: syq at debian dot org Target Milestone: --- https://sources.debian.org/patches/binutils/2.33.1-2/mips64-default-n64.diff/ For historical reasons, mips64*-linux-gnu is set as N32, so some distributions like Debian use -gnuabi64 as their ABI section for N64 port. So we should set all mips*64*-linux-gnuabi64 as N64. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/25136] mips64: set abi as N64 when the ABI section triple is gnuabi64
https://sourceware.org/bugzilla/show_bug.cgi?id=25136 --- Comment #1 from YunQiang Su --- The reason change mips64* to mips*64* is that for MIPS r6, we use something like mipsisa64r6el-linux-gnuabi64 Since config.guess and config.sub use it. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/25135] Cannot link because of ".dynsym local symbol at index"
https://sourceware.org/bugzilla/show_bug.cgi?id=25135 Alan Modra changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |INVALID --- Comment #3 from Alan Modra --- I fail to see what you're complaining about then. The message doesn't say "warning", but that is what it is when reported for a shared library. Linking should complete successfully despite the warning unless you pass --fatal-warnings to the linker. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/25135] Cannot link because of ".dynsym local symbol at index"
https://sourceware.org/bugzilla/show_bug.cgi?id=25135 Benjamin changed: What|Removed |Added Resolution|INVALID |FIXED --- Comment #4 from Benjamin --- (In reply to Alan Modra from comment #3) > I fail to see what you're complaining about then. The message doesn't say > "warning", but that is what it is when reported for a shared library. > Linking should complete successfully despite the warning unless you pass > --fatal-warnings to the linker. After further inspection the linking does complete and g++ output status is 0. For some reason it is the tools that I am using (IDE and others) that understand this as an error (not warning) and report a build failure... Thank you for your time. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/25135] Cannot link because of ".dynsym local symbol at index"
https://sourceware.org/bugzilla/show_bug.cgi?id=25135 --- Comment #5 from Benjamin --- Would it be possible to add "warning:" to the error message? Pretty much like what is done with the lld messages. Looking at your code base it seems that it is what is usually done and it would allow external tools to detect it as warnings, not errors. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/25135] Cannot link because of ".dynsym local symbol at index"
https://sourceware.org/bugzilla/show_bug.cgi?id=25135 --- Comment #6 from Alan Modra --- There are many warnings that don't contain the word "warning". There are even errors that contain the word "warning". If you would like to correct this, please audit the source for this sort of problem and submit patches to binut...@sourceware.org. A patch just fixing this one warning isn't very useful. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/25138] Incorrect subtraction in linker script
https://sourceware.org/bugzilla/show_bug.cgi?id=25138 Alan Modra changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED CC||amodra at gmail dot com Resolution|--- |WORKSFORME --- Comment #1 from Alan Modra --- You didn't show much of your linker script but I suspect you defined _nvram_data, _envram, and _nvram_data_size inside some output section, in which case all the values are zero *relative to that output section*. If that is so, the value shown by the 2.33 map file for _nvram_data_size is correct and will match that reported by nm for the symbol. (The patch you identify fixed bugs in symbol values shown in map files.) Assign _nvram_data_size after the output section, and of course not inside some other output section. -- You are receiving this mail because: You are on the CC list for the bug.
