This is the mail archive of the binutils@sourceware.org mailing list for the binutils project.
| Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
|---|---|---|
| Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
| Other format: | [Raw text] | |
Hi H.J.According to the logic, if there are versioned symbols with the same symbol name and version, e.g. sym_name@VERSION. It will treated as multiple definition of symbol. Until now, the behavior is the same as before. The linker will throw multiple definition error.
With the change here, and with "--allow-multiple-definition" option provided, the first defined one will fully override a later definition. For default version symbol, it's slightly different.You can only declare one version of a symbol as the default in this manner; otherwise you would effectively have multiple definitions of the same symbol.
For sym_name@@VERSION, two symbols will be added. One is sym_name@@VERSION. As above, you cannot define sym_name@@VERSION twice. what's more, _bfd_elf_add_default_symbol will add one "sym_name" indirect symbol. The check for this symbol definition is exclude explicitly in this patch as the condition indicates.
pr21703-3.s
.text
.global foo
.symver foo, foo@FOO
.type foo, %function
foo:
.space 4
.size foo, 4
pr21703-4.s
.text
.global bar
.symver bar, foo@FOO
.type bar, %function
bar:
.space 16
.size bar, 16
With the following command line:
as-new pr21703-3.o pr21703-3.s
as-new pr21703-4.o pr21703-4.s
ld-new -o pr21703 -z norelro pr21703-3.o pr21703-4.o --allow-multiple-definition
without the patch,
The symbol table of final object is:
Num: Value Size Type Bind Vis Ndx Name
0: 0000000000000000 0 NOTYPE LOCAL DEFAULT UND
1: 0000000000400078 0 SECTION LOCAL DEFAULT 1
2: 0000000000600090 0 OBJECT LOCAL DEFAULT 1 _GLOBAL_OFFSET_TABLE_
3: 0000000000400078 16 FUNC GLOBAL DEFAULT 1 foo@FOO
4: 000000000060008c 0 NOTYPE GLOBAL DEFAULT 1 __bss_start
5: 000000000060008c 0 NOTYPE GLOBAL DEFAULT 1 _edata
6: 0000000000600090 0 NOTYPE GLOBAL DEFAULT 1 _end
7: 000000000040007c 16 FUNC GLOBAL DEFAULT 1 bar
foo@FOO is mapped to foo (the value) in pr21703-3.o with size from bar in pr21703-4.o.
with the patch:
Num: Value Size Type Bind Vis Ndx Name
0: 0000000000000000 0 NOTYPE LOCAL DEFAULT UND
1: 0000000000400078 0 SECTION LOCAL DEFAULT 1
2: 0000000000600090 0 OBJECT LOCAL DEFAULT 1 _GLOBAL_OFFSET_TABLE_
3: 0000000000400078 4 FUNC GLOBAL DEFAULT 1 foo@FOO
4: 000000000060008c 0 NOTYPE GLOBAL DEFAULT 1 __bss_start
5: 000000000060008c 0 NOTYPE GLOBAL DEFAULT 1 _edata
6: 0000000000600090 0 NOTYPE GLOBAL DEFAULT 1 _end
7: 000000000040007c 16 FUNC GLOBAL DEFAULT 1 bar
foo@FOO is fully mapped to foo.
The above behavior is observed both on x64 and arm.
Is this the kind of test you are suggesting?
Regards,
Renlin
On 11/10/17 22:24, H.J. Lu wrote:
On 10/11/17, Renlin Li <renlin.li@foss.arm.com> wrote:Ping ~ Regards, Renlin On 12/07/17 15:21, Renlin Li wrote:Hi, On 12/07/17 15:17, Renlin Li wrote:Hi H.J, On 11/07/17 19:04, H.J. Lu wrote:On Tue, Jul 11, 2017 at 10:31 AM, Renlin Li <renlin.li@foss.arm.com> wrote:ld/ChangeLog: 2017-07-11 Renlin Li <renlin.li@arm.com> PR ld/21703 * testsuite/ld-elf/elf.exp : Run new test case. * testsuite/ld-elf/pr21703-1.s: New. * testsuite/ld-elf/pr21703-2.s: New. * testsuite/ld-elf/pr21703.sd: New. bfd/ChangeLog: 2017-07-11 Renlin Li <renlin.li@arm.com> PR ld/21703 * elflink.c (_bfd_elf_merge_symbol): Handle multiple symbol definition case.Please add some a test to cover versioned_sym in:This is because _bfd_elf_merge_symbol will be called for versioned symbol with the short name. For example, ld/testsuite/ld-elf/pr18735.s It defines foo and an alias foo@FOO. in addition to that, _bfd_elf_add_default_symbol will be called for foo@FOO to create a new symbol which short name (foo in this case). For the original condition, this will generate a multiple definition error. "matched" flag seems introduced by you. Here I use it to differentiate the versioned symbol case from normal symbol definition./* There are multiple definitions of a normal symbol. */ + if (olddef && !olddyn && !oldweak && newdef && !newdyn && !newweak + && !versioned_sym && h->def_regular) + { What should happen if symbol is defined in linker script or command line?This is a good point. I missed this case. I have read the code a little bit. Multiple definition handling is broken for this case. the following command throws error about multiple definitions. ld --defsym foo=0x10 pr21703-1.o -o tmpdir/dump The command links fine, generating foo symbol with *MERGED* information from two definition. ld pr21703-1.o --defsym foo=0x10 -o tmpdir/dump 00000010 4 OBJECT GLOBAL DEFAULT ABS foo The functions to process symbol definitions from object file and command line is different. Multiple-definition is not handled for command line symbol definition. And the symbol value is overwritten by the symbol absolute value. The definition section is overwritten as well to ABS section. It's intended. Alan had a patch to allow multiple symbol definition without provide --allow-multiple-definition option. https://sourceware.org/ml/binutils/2014-12/msg00247.html The partial symbol information overwritten is the same problem as the original bug I reported.So there is another question. What's the expected override behavior for command line, linker script and object/(static/dynamic)library defined symbols? Is there a defined rule for those cases that I have missed? Thanks! Renlin.With or without the patch, the behavior remains the same. The symbol defined in command line option don't set h->def_regular flag. h->ldscript_def flag is set instead. Regards, RenlinI still like to see testcases with symbol versioning and --allow-multiple-definition.
| Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
|---|---|---|
| Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |