nm: add some .text.* support for symbols

.text.* are generated when using -ffunction-sections.

makes nm(1) to correctly identify the function symbols as N_TEXT.

found while debugging librsvg build error, where some symbols where not found by libtool(1) (which is using nm(1)).

ok gkoehler@

fix typo; from Kris Katterjohn

Delete unused phdr functions; move declarations for functions internal to
elf.c to that file

ok mpi@

recognize .openbsd.randomdata section and indicate it roughly, so
that objects within it are identified as being in read-only space.
ok guenther

Use <elf.h> instead of <elf_abi.h>

ok jasper@, jca@, deraadt@

Remove NULL-checks before free(). ok tb@

Grow -P (POSIX output) and -t (POSIX output radix) support. Several 3rd-party
software depend upon this.

Requested by feinerer@, ok millert@; manpage bits ok jmc@

This patch ensure that when an error is detected, the freed variables in
elf_symloadx() are reinitialised.

Else show_file() in nm.c will used these variables, even if they has
just been freed. (nm.c +689).

Problem found by afl.

ok miod@

This patch ensure that e_shentsize (sections header's size in bytes) is
big enough to fill at least one Elf_Shdr.

While here, inverts calloc() arguments to be calloc(nmemb, size),
according to fread() call after.

This problem was found with afl, with e_shentsize=1.

ok miod@

corrects a read after bound that occurs in strcmp (line just
after the added bound check).

Found with afl.

ok miod@

add some check before accessing data:
- if section header table is be present
- consistency of section header table size

ok miod@

Add -D option for displaying the dynamic symbol table

ok miod@

ELF weak symbols are _not_ like a.out indirect symbols. Nothing in ELF is,
so remove N_INDR handling.

ok kettenis@

SIZE_MAX is standard, we should be using it in preference to the
obsolete SIZE_T_MAX. OK miod@ beck@

Missing initialization; tjenahej@speedmail.se

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

unsigned char casts for ctype
ok krw

proto and such cleanup
ok guenther

Correctly report `weak' status of non-function symbols. Makes nm agree with
objdump -t on, say, _DYNAMIC in libpthread.

Support symbols in .tbss and .tdata ELF sections

With this change nm(1) prints the expected symbol type ('B' or 'D')
for variables in the .tbss and .tdata sections respectively, instead
of '?'. This would be what binutils do for non-automatic variables
that are marked as thread-local with the __thread attribute in GCC.

help and ok miod@, guenther@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

use calloc() to avoid malloc(n * m) overflows; checked by djm canacar jsg

Fix a NULL dereference if the symbol table size is zero; ok millert@ deraadt@

Do all allocation and deallocation of shstr in elf_symload() to
prevent double-freeing shstr.

Also don't free known NULL values (*pnames and *psnames).

Initially inspired by Charles Longeau's first diff to tech@; he
later sent a second diff to tech@ nearly identical to this commit.
Freaky.

OK mickey@.

check st_name validity also check for overflow and not only for zero; found by Charles Longeau <chl@tuxfamily.org>

Some fixes for running nm on ELF binaries.
- Fix printing the symbol type of undefined symbols.
- Fall back onto the dynamic symbol table if the normal symbol table is not
available (i.e. if the binary has been stripped).

ok mickey@

Add elf{32,64}_load_phdrs() for loading program headers.

ok mickey@

add rcisd[]

parse the cpu-specific ids based on the e_machine

move out elf code into elf.c and generate 32 and 64 versions of it;
allow operation on size-mixed host-target as well as ensianess for
both a.out and elf (though a.out 64bit probably does not exist ;)
tested on hppa-sparc64-i386-vax intermixen

factor out code for mapping section index into nlist type

grok STT_SECTION and beautify STT_OBJECT

STT_FUNC can be an absolute too

dig .plt*, .got*, .init, .fini sections

no need to force N_EXT as deducted later from type

there is more than one way to define symbol type in elf...; kill a debugging printf too

translate names in .sbss section as static N_BSS; found by drahn@

kill unused local vars; from deraadt@

nm digs elf and merges w/ size; millert@ ok

fix typo; from Kris Katterjohn

Delete unused phdr functions; move declarations for functions internal to
elf.c to that file

ok mpi@

recognize .openbsd.randomdata section and indicate it roughly, so
that objects within it are identified as being in read-only space.
ok guenther

Use <elf.h> instead of <elf_abi.h>

ok jasper@, jca@, deraadt@

Remove NULL-checks before free(). ok tb@

Grow -P (POSIX output) and -t (POSIX output radix) support. Several 3rd-party
software depend upon this.

Requested by feinerer@, ok millert@; manpage bits ok jmc@

This patch ensure that when an error is detected, the freed variables in
elf_symloadx() are reinitialised.

Else show_file() in nm.c will used these variables, even if they has
just been freed. (nm.c +689).

Problem found by afl.

ok miod@

This patch ensure that e_shentsize (sections header's size in bytes) is
big enough to fill at least one Elf_Shdr.

While here, inverts calloc() arguments to be calloc(nmemb, size),
according to fread() call after.

This problem was found with afl, with e_shentsize=1.

ok miod@

corrects a read after bound that occurs in strcmp (line just
after the added bound check).

Found with afl.

ok miod@

add some check before accessing data:
- if section header table is be present
- consistency of section header table size

ok miod@

Add -D option for displaying the dynamic symbol table

ok miod@

ELF weak symbols are _not_ like a.out indirect symbols. Nothing in ELF is,
so remove N_INDR handling.

ok kettenis@

SIZE_MAX is standard, we should be using it in preference to the
obsolete SIZE_T_MAX. OK miod@ beck@

Missing initialization; tjenahej@speedmail.se

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

unsigned char casts for ctype
ok krw

proto and such cleanup
ok guenther

Correctly report `weak' status of non-function symbols. Makes nm agree with
objdump -t on, say, _DYNAMIC in libpthread.

Support symbols in .tbss and .tdata ELF sections

With this change nm(1) prints the expected symbol type ('B' or 'D')
for variables in the .tbss and .tdata sections respectively, instead
of '?'. This would be what binutils do for non-automatic variables
that are marked as thread-local with the __thread attribute in GCC.

help and ok miod@, guenther@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

use calloc() to avoid malloc(n * m) overflows; checked by djm canacar jsg

Fix a NULL dereference if the symbol table size is zero; ok millert@ deraadt@

Do all allocation and deallocation of shstr in elf_symload() to
prevent double-freeing shstr.

Also don't free known NULL values (*pnames and *psnames).

Initially inspired by Charles Longeau's first diff to tech@; he
later sent a second diff to tech@ nearly identical to this commit.
Freaky.

OK mickey@.

check st_name validity also check for overflow and not only for zero; found by Charles Longeau <chl@tuxfamily.org>

Some fixes for running nm on ELF binaries.
- Fix printing the symbol type of undefined symbols.
- Fall back onto the dynamic symbol table if the normal symbol table is not
available (i.e. if the binary has been stripped).

ok mickey@

Add elf{32,64}_load_phdrs() for loading program headers.

ok mickey@

add rcisd[]

parse the cpu-specific ids based on the e_machine

move out elf code into elf.c and generate 32 and 64 versions of it;
allow operation on size-mixed host-target as well as ensianess for
both a.out and elf (though a.out 64bit probably does not exist ;)
tested on hppa-sparc64-i386-vax intermixen

factor out code for mapping section index into nlist type

grok STT_SECTION and beautify STT_OBJECT

STT_FUNC can be an absolute too

dig .plt*, .got*, .init, .fini sections

no need to force N_EXT as deducted later from type

there is more than one way to define symbol type in elf...; kill a debugging printf too

translate names in .sbss section as static N_BSS; found by drahn@

kill unused local vars; from deraadt@

nm digs elf and merges w/ size; millert@ ok

Delete unused phdr functions; move declarations for functions internal to
elf.c to that file

ok mpi@

recognize .openbsd.randomdata section and indicate it roughly, so
that objects within it are identified as being in read-only space.
ok guenther

Use <elf.h> instead of <elf_abi.h>

ok jasper@, jca@, deraadt@

Remove NULL-checks before free(). ok tb@

Grow -P (POSIX output) and -t (POSIX output radix) support. Several 3rd-party
software depend upon this.

Requested by feinerer@, ok millert@; manpage bits ok jmc@

This patch ensure that when an error is detected, the freed variables in
elf_symloadx() are reinitialised.

Else show_file() in nm.c will used these variables, even if they has
just been freed. (nm.c +689).

Problem found by afl.

ok miod@

This patch ensure that e_shentsize (sections header's size in bytes) is
big enough to fill at least one Elf_Shdr.

While here, inverts calloc() arguments to be calloc(nmemb, size),
according to fread() call after.

This problem was found with afl, with e_shentsize=1.

ok miod@

corrects a read after bound that occurs in strcmp (line just
after the added bound check).

Found with afl.

ok miod@

add some check before accessing data:
- if section header table is be present
- consistency of section header table size

ok miod@

Add -D option for displaying the dynamic symbol table

ok miod@

ELF weak symbols are _not_ like a.out indirect symbols. Nothing in ELF is,
so remove N_INDR handling.

ok kettenis@

SIZE_MAX is standard, we should be using it in preference to the
obsolete SIZE_T_MAX. OK miod@ beck@

Missing initialization; tjenahej@speedmail.se

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

unsigned char casts for ctype
ok krw

proto and such cleanup
ok guenther

Correctly report `weak' status of non-function symbols. Makes nm agree with
objdump -t on, say, _DYNAMIC in libpthread.

Support symbols in .tbss and .tdata ELF sections

With this change nm(1) prints the expected symbol type ('B' or 'D')
for variables in the .tbss and .tdata sections respectively, instead
of '?'. This would be what binutils do for non-automatic variables
that are marked as thread-local with the __thread attribute in GCC.

help and ok miod@, guenther@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

use calloc() to avoid malloc(n * m) overflows; checked by djm canacar jsg

Fix a NULL dereference if the symbol table size is zero; ok millert@ deraadt@

Do all allocation and deallocation of shstr in elf_symload() to
prevent double-freeing shstr.

Also don't free known NULL values (*pnames and *psnames).

Initially inspired by Charles Longeau's first diff to tech@; he
later sent a second diff to tech@ nearly identical to this commit.
Freaky.

OK mickey@.

check st_name validity also check for overflow and not only for zero; found by Charles Longeau <chl@tuxfamily.org>

Some fixes for running nm on ELF binaries.
- Fix printing the symbol type of undefined symbols.
- Fall back onto the dynamic symbol table if the normal symbol table is not
available (i.e. if the binary has been stripped).

ok mickey@

Add elf{32,64}_load_phdrs() for loading program headers.

ok mickey@

add rcisd[]

parse the cpu-specific ids based on the e_machine

move out elf code into elf.c and generate 32 and 64 versions of it;
allow operation on size-mixed host-target as well as ensianess for
both a.out and elf (though a.out 64bit probably does not exist ;)
tested on hppa-sparc64-i386-vax intermixen

factor out code for mapping section index into nlist type

grok STT_SECTION and beautify STT_OBJECT

STT_FUNC can be an absolute too

dig .plt*, .got*, .init, .fini sections

no need to force N_EXT as deducted later from type

there is more than one way to define symbol type in elf...; kill a debugging printf too

translate names in .sbss section as static N_BSS; found by drahn@

kill unused local vars; from deraadt@

nm digs elf and merges w/ size; millert@ ok

recognize .openbsd.randomdata section and indicate it roughly, so
that objects within it are identified as being in read-only space.
ok guenther

Use <elf.h> instead of <elf_abi.h>

ok jasper@, jca@, deraadt@

Remove NULL-checks before free(). ok tb@

Grow -P (POSIX output) and -t (POSIX output radix) support. Several 3rd-party
software depend upon this.

Requested by feinerer@, ok millert@; manpage bits ok jmc@

This patch ensure that when an error is detected, the freed variables in
elf_symloadx() are reinitialised.

Else show_file() in nm.c will used these variables, even if they has
just been freed. (nm.c +689).

Problem found by afl.

ok miod@

This patch ensure that e_shentsize (sections header's size in bytes) is
big enough to fill at least one Elf_Shdr.

While here, inverts calloc() arguments to be calloc(nmemb, size),
according to fread() call after.

This problem was found with afl, with e_shentsize=1.

ok miod@

corrects a read after bound that occurs in strcmp (line just
after the added bound check).

Found with afl.

ok miod@

add some check before accessing data:
- if section header table is be present
- consistency of section header table size

ok miod@

Add -D option for displaying the dynamic symbol table

ok miod@

ELF weak symbols are _not_ like a.out indirect symbols. Nothing in ELF is,
so remove N_INDR handling.

ok kettenis@

SIZE_MAX is standard, we should be using it in preference to the
obsolete SIZE_T_MAX. OK miod@ beck@

Missing initialization; tjenahej@speedmail.se

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

unsigned char casts for ctype
ok krw

proto and such cleanup
ok guenther

Correctly report `weak' status of non-function symbols. Makes nm agree with
objdump -t on, say, _DYNAMIC in libpthread.

Support symbols in .tbss and .tdata ELF sections

With this change nm(1) prints the expected symbol type ('B' or 'D')
for variables in the .tbss and .tdata sections respectively, instead
of '?'. This would be what binutils do for non-automatic variables
that are marked as thread-local with the __thread attribute in GCC.

help and ok miod@, guenther@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

use calloc() to avoid malloc(n * m) overflows; checked by djm canacar jsg

Fix a NULL dereference if the symbol table size is zero; ok millert@ deraadt@

Do all allocation and deallocation of shstr in elf_symload() to
prevent double-freeing shstr.

Also don't free known NULL values (*pnames and *psnames).

Initially inspired by Charles Longeau's first diff to tech@; he
later sent a second diff to tech@ nearly identical to this commit.
Freaky.

OK mickey@.

check st_name validity also check for overflow and not only for zero; found by Charles Longeau <chl@tuxfamily.org>

Some fixes for running nm on ELF binaries.
- Fix printing the symbol type of undefined symbols.
- Fall back onto the dynamic symbol table if the normal symbol table is not
available (i.e. if the binary has been stripped).

ok mickey@

Add elf{32,64}_load_phdrs() for loading program headers.

ok mickey@

add rcisd[]

parse the cpu-specific ids based on the e_machine

move out elf code into elf.c and generate 32 and 64 versions of it;
allow operation on size-mixed host-target as well as ensianess for
both a.out and elf (though a.out 64bit probably does not exist ;)
tested on hppa-sparc64-i386-vax intermixen

factor out code for mapping section index into nlist type

grok STT_SECTION and beautify STT_OBJECT

STT_FUNC can be an absolute too

dig .plt*, .got*, .init, .fini sections

no need to force N_EXT as deducted later from type

there is more than one way to define symbol type in elf...; kill a debugging printf too

translate names in .sbss section as static N_BSS; found by drahn@

kill unused local vars; from deraadt@

nm digs elf and merges w/ size; millert@ ok