modpost: various fixes, cleanups, optimizations
cover modpost: various fixes, cleanups, optimizations
Masahiro Yamada (29):
modpost: drop RCS/CVS $Revision handling in MODULE_VERSION()
modpost: do not call get_modinfo() for vmlinux
modpost: add read_text_file() and get_line() helpers
modpost: fix potential mmap'ed file overrun in get_src_version()
modpost: re-add warning about missing *.mod file
modpost: avoid false-positive file open error
modpost: use read_text_file() and get_line() for reading text files
modpost: remove get_next_text() and make {grab,release_}file static
kbuild: disallow multi-word in M= or KBUILD_EXTMOD
modpost: move -T option close to the modpost command
modpost: pass -N option only for modules modpost
modpost: move external module options
modpost: load KBUILD_EXTRA_SYMBOLS files in order
modpost: track if the symbol origin is a dump file or ELF object
modpost: allow to pass -i option multiple times remove -e option
modpost: rename ext_sym_list to dump_list
modpost: re-add -e to set external_module flag
modpost: show warning if vmlinux is not found when processing modules
modpost: show warning if it fails to read symbol dump file
modpost: generate vmlinux.symvers and reuse it for the second modpost
modpost: remove -s option
modpost: remove mod->is_dot_o struct member
modpost: remove is_vmlinux() call in check_for_{gpl_usage,unused}()
modpost: add mod->is_vmlinux struct member
modpost: remove mod->skip struct member
modpost: set have_vmlinux in new_module()
modpost: strip .o from modname before calling new_module()
modpost: remove is_vmlinux() helper
modpost: change elf_info->size to size_t
.gitignore | 1 + Makefile | 15 +- scripts/Makefile.modpost | 29 ++-- scripts/link-vmlinux.sh | 2 - scripts/mod/modpost.c | 309 ++++++++++++++++++--------------------- scripts/mod/modpost.h | 17 +-- scripts/mod/sumversion.c | 114 ++++----------- 7 files changed, 195 insertions(+), 292 deletions(-)
Signature · 1 line
2.25.1
PATCH 1/29 modpost: drop RCS/CVS $Revision handling in MODULE_VERSION()
As far as I understood, this code gets rid of '$Revision$' or '$Revision:' of CVS, RCS or whatever in MODULE_VERSION() tags.
Remove the primeval code.
---
scripts/mod/modpost.c | 3 -- scripts/mod/modpost.h | 4 --- scripts/mod/sumversion.c | 66 ---------------------------------------- 3 files changed, 73 deletions(-)
scripts/mod/modpost.c +0 -3
@@ -2074,9 +2074,6 @@ static void read_symbols(const char *modname)check_sec_ref (mod, modname, &info); version =get_modinfo (&info,"version" ); -if (version) -maybe_frob_rcs_version (modname, version, info.modinfo, - version - (char *)info.hdr);if (version || (all_versions && !is_vmlinux (modname)))get_src_version (modname, mod->srcversion,sizeof (mod->srcversion)-1 );
scripts/mod/modpost.h +0 -4
@@ -187,10 +187,6 @@ void handle_moddevtable(struct module *mod, struct elf_info *info,void add_moddevtable (struct buffer *buf,struct module *mod);/* sumversion.c */ -void maybe_frob_rcs_version (const char *modfilename, -char *version, -void *modinfo, -unsigned long modinfo_offset);void get_src_version (const char *modname,char sum[],unsigned sumlen);/* from modpost.c */
scripts/mod/sumversion.c +0 -66
69 lines not on this page · raw message
Signature · 1 line
2.25.1
PATCH 2/29 modpost: do not call get_modinfo() for vmlinux
In reply to Masahiro YamadaThe three calls of get_modinfo() ("license", "import_ns", "version") always return NULL for vmlinux because the built-in module info is prefixed with __MODULE_INFO_PREFIX.
It is harmless to call get_modinfo(), but there is no point to search for what apparently does not exist.
---
scripts/mod/modpost.c | 45 +++++++++++++++++++++++-------------------- 1 file changed, 24 insertions(+), 21 deletions(-)
scripts/mod/modpost.c +24 -21
@@ -2014,25 +2014,26 @@ static void read_symbols(const char *modname) mod->skip =1 ; } - license =get_modinfo (&info,"license" ); -if (!license && !is_vmlinux (modname)) -warn ("missing MODULE_LICENSE() in %s\n" -"see include/linux/module.h for " -"more information\n" , modname); -while (license) { -if (license_is_gpl_compatible (license)) - mod->gpl_compatible =1 ; -else { - mod->gpl_compatible =0 ; -break ; +if (!is_vmlinux (modname)) { + license =get_modinfo (&info,"license" ); +if (!license) +warn ("missing MODULE_LICENSE() in %s\n" , modname); +while (license) { +if (license_is_gpl_compatible (license)) + mod->gpl_compatible =1 ; +else { + mod->gpl_compatible =0 ; +break ; + } + license =get_next_modinfo (&info,"license" , license); } - license =get_next_modinfo (&info,"license" , license); - } - namespace =get_modinfo (&info,"import_ns" ); -while (namespace) { -add_namespace (&mod->imported_namespaces, namespace); - namespace =get_next_modinfo (&info,"import_ns" , namespace); + namespace =get_modinfo (&info,"import_ns" ); +while (namespace) { +add_namespace (&mod->imported_namespaces, namespace); + namespace =get_next_modinfo (&info,"import_ns" , + namespace); + } }for (sym = info.symtab_start; sym < info.symtab_stop; sym++) {
@@ -2073,10 +2074,12 @@ static void read_symbols(const char *modname)if (!is_vmlinux (modname) || vmlinux_section_warnings)check_sec_ref (mod, modname, &info); - version =get_modinfo (&info,"version" ); -if (version || (all_versions && !is_vmlinux (modname))) -get_src_version (modname, mod->srcversion, -sizeof (mod->srcversion)-1 ); +if (!is_vmlinux (modname)) { + version =get_modinfo (&info,"version" ); +if (version || all_versions) +get_src_version (modname, mod->srcversion, +sizeof (mod->srcversion) -1 ); + }parse_elf_finish (&info);
Signature · 1 line
2.25.1
PATCH 3/29 modpost: add read_text_file() and get_line() helpers
In reply to Masahiro Yamadamodpost uses grab_file() to open a file, but it is not suitable for a text file because the mmaped file is not terminated by null byte. Actually, I see some issues for the use of grab_file().
The new helper, read_text_file() loads the whole file content into a malloc'ed buffer, and appends a null byte. Then, get_line() reads each line.
To handle text files, I intend to replace as follows:
grab_file() -> read_text_file()
get_new_line() -> get_line()
---
scripts/mod/modpost.c | 36 ++++++++++++++++++++++++++++++++++++ scripts/mod/modpost.h | 2 ++ 2 files changed, 38 insertions(+)
scripts/mod/modpost.c +36 -0
@@ -112,6 +112,42 @@ void *do_nofail(void *ptr, const char *expr)return ptr; } +char *read_text_file (const char *filename) +{ +struct stat st; +int fd; +char *buf; + + fd =open (filename,O_RDONLY ); +if (fd <0 ) +return NULL ; + +if (fstat (fd, &st) <0 ) +return NULL ; + + buf =NOFAIL (malloc (st.st_size +1 )); + +if (read (fd, buf, st.st_size) != st.st_size) { +free (buf); + buf =NULL ; + goto close; + } + buf[st.st_size] ='\0' ; +close: +close (fd); + +return buf; +} + +char *get_line (char **stringp) +{ +/* do not return the unwanted extra line at EOF */ +if (*stringp && **stringp =='\0' ) +return NULL ; + +return strsep (stringp,"\n" ); +} +/* A list of all modules we processed */ static struct module *modules;
scripts/mod/modpost.h +2 -0
@@ -190,6 +190,8 @@ void add_moddevtable(struct buffer *buf, struct module *mod);void get_src_version (const char *modname,char sum[],unsigned sumlen);/* from modpost.c */ +char *read_text_file (const char *filename); +char *get_line (char **stringp);void *grab_file (const char *filename,unsigned long *size);char *get_next_line (unsigned long *pos,void *file,unsigned long size);void release_file (void *file,unsigned long size);
Signature · 1 line
2.25.1
Re: [PATCH 03/29] modpost: add read_text_file() and get_line() helpers
Quoted 6 lines · Masahiro Yamada
scripts/mod/modpost.c +6
-0+char *read_text_file (const char *filename) +{ +struct stat st; +int fd; +char *buf; +
scripts/mod/modpost.c void *do_nofail(void *ptr, const char *expr) · line 130
Quoted 10 lines
scripts/mod/modpost.c +10
-0+ fd =open (filename,O_RDONLY ); +if (fd <0 ) +return NULL ; + +if (fstat (fd, &st) <0 ) +return NULL ; + + buf =NOFAIL (malloc (st.st_size +1 )); + +if (read (fd, buf, st.st_size) != st.st_size) {
Is this sensible coding ? I've always been taught read() can return early/short for a number of reasons and we must not assume this is an error.
The 'normal' way to read a file is something like:
for (;;) {
ssize_t ret = read(fd, buf + size, st.st_size - size);
if (ret < 0) {
free(buf);
buf = NULL;
goto close;
}
if (!ret)
break;
size += ret;
}Quoted 10 lines
scripts/mod/modpost.c +10
-0+free (buf); + buf =NULL ; + goto close; + } + buf[st.st_size] ='\0' ; +close: +close (fd); + +return buf; +}
Re: [PATCH 03/29] modpost: add read_text_file() and get_line() helpers
Quoted 39 lines · Peter Zijlstra
Quoted 6 lines · Masahiro Yamada
scripts/mod/modpost.c +6
-0+char *read_text_file (const char *filename) +{ +struct stat st; +int fd; +char *buf; +scripts/mod/modpost.c void *do_nofail(void *ptr, const char *expr) · line 130
Quoted 10 lines
scripts/mod/modpost.c +10
-0+ fd =open (filename,O_RDONLY ); +if (fd <0 ) +return NULL ; + +if (fstat (fd, &st) <0 ) +return NULL ; + + buf =NOFAIL (malloc (st.st_size +1 )); + +if (read (fd, buf, st.st_size) != st.st_size) {Is this sensible coding ? I've always been taught
read()can return early/short for a number of reasons and we must not assume this is an error.The 'normal' way to read a file is something like:
for (;;) { ssize_t ret = read(fd, buf + size, st.st_size - size); if (ret < 0) { free(buf); buf = NULL; goto close; } if (!ret) break; size += ret; }
Quoted 10 lines
scripts/mod/modpost.c void *do_nofail(void *ptr, const char *expr) · line 140
Quoted 10 lines
scripts/mod/modpost.c +10
-0+free (buf); + buf =NULL ; + goto close; + } + buf[st.st_size] ='\0' ; +close: +close (fd); + +return buf; +}
In theory, I think yes.
But, is it necessary when we know
it is reading a regular file?
The specification [1] says this:
"The value returned may be less than nbyte if the number of bytes left in the file is less than nbyte, if the read() request was interrupted by a signal, or if the file is a pipe or FIFO or special file and has fewer than nbyte bytes immediately available for reading."
This case does not meet any of 'if ...' parts.
[1] https://pubs.opengroup.org/onlinepubs/000095399/functions/read.html
Signature · 2 lines
Best Regards Masahiro Yamada
Re: [PATCH 03/29] modpost: add read_text_file() and get_line() helpers
Quoted 10 lines · Masahiro Yamada
The specification [1] says this:
"The value returned may be less than nbyte if the number of bytes left in the file is less than nbyte, if the
read()request was interrupted by a signal, or if the file is a pipe or FIFO or special file and has fewer than nbyte bytes immediately available for reading."This case does not meet any of 'if ...' parts.
So nobody ever ^Z's their build? I really don't think you can assume that you'll never get a signal. That's just asking for trouble.
Doing the right thing here is 'trivial', there is absolutely no reason to not do it.
Re: [PATCH 03/29] modpost: add read_text_file() and get_line() helpers
Quoted 12 lines · Peter Zijlstra
Quoted 9 lines · Masahiro Yamada
The specification [1] says this:
"The value returned may be less than nbyte if the number of bytes left in the file is less than nbyte, if the
read()request was interrupted by a signal, or if the file is a pipe or FIFO or special file and has fewer than nbyte bytes immediately available for reading."
Quoted 7 lines
Quoted 1 line
This case does not meet any of 'if ...' parts.
So nobody ever ^Z's their build? I really don't think you can assume that you'll never get a signal. That's just asking for trouble.
Doing the right thing here is 'trivial', there is absolutely no reason to not do it.
In my testing, read() seems robust against ^Z,
but perhaps it might be implementation-defined ?
I think you are right.
We can gain extra safety with a trivial change.
--
Best Regards
Masahiro Yamada
PATCH 4/29 modpost: fix potential mmap'ed file overrun in get_src_version()
In reply to Masahiro YamadaI do not know how reliably this function works, but it seems dangerous to me, at least.
The function call
strchr(sources, '\n');
... continues searching until it finds '\n' or it reaches the '\0' terminator. In other words, 'sources' should be a null-terminated string.
However, grab_file() just mmaps a file, so 'sources' is not terminated with null byte. If the file does not contain '\n' at all, strchr() will go beyond the mmap'ed memory.
Instead, use read_text_file(), which loads the file content into a malloc'ed buffer, appending null-byte.
Here we are interested only in the first line of *.mod files. Use get_line() helper to get the first line.
---
scripts/mod/sumversion.c | 27 ++++++++++++--------------- 1 file changed, 12 insertions(+), 15 deletions(-)
scripts/mod/sumversion.c +12 -15
@@ -392,40 +392,37 @@ static int parse_source_files(const char *objfile, struct md4_ctx *md)/* Calc and record src checksum. */ void get_src_version (const char *modname,char sum[],unsigned sumlen) { -void *file; -unsigned long len; +char *buf, *pos, *firstline;struct md4_ctx md; -char *sources, *end, *fname; +char *fname;char filelist[PATH_MAX +1 ];/* objects for a module are listed in the first line of *.mod file. */ snprintf (filelist,sizeof (filelist),"%.*smod" , (int )strlen (modname) -1 , modname); - file =grab_file (filelist, &len); -if (!file) + buf =read_text_file (filelist); +if (!buf)/* not a module or .mod file missing - ignore */ return ; - sources = file; - - end =strchr (sources,'\n' ); -if (!end) { + pos = buf; + firstline =get_line (&pos); +if (!firstline) {warn ("bad ending versions file for %s\n" , modname); - goto release; + goto free; } - *end ='\0' ;md4_init (&md); -while ((fname =strsep (&sources," " )) !=NULL ) { +while ((fname =strsep (&firstline," " ))) {if (!*fname)continue ;if (!(is_static_library (fname)) && !parse_source_files (fname, &md)) - goto release; + goto free; }md4_final_ascii (&md, sum, sumlen); -release: -release_file (file, len); +free: +free (buf); }
Signature · 1 line
2.25.1
PATCH 5/29 modpost: re-add warning about missing *.mod file
In reply to Masahiro YamadaThis reverts 4be40e22233c ("kbuild: do not emit src version warning for non-modules").
I do not fully understand what that commit addressed, but commit 91341d4b2c19 ("kbuild: introduce new option to enhance section mismatch analysis") introduced partial section checks by using modpost. built-in.o was parsed by modpost. Even modules had a problem because *.mod files were created after the modpost check.
Commit b7dca6dd1e59 ("kbuild: create *.mod with full directory path and remove MODVERDIR") stopped doing that. Now that modpost is only invoked after the directory descend, *.mod files should always exist at the modpost stage. If not, something went wrong, which should be warned.
---
scripts/mod/sumversion.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
scripts/mod/sumversion.c +4 -2
@@ -402,9 +402,11 @@ void get_src_version(const char *modname, char sum[], unsigned sumlen) (int)strlen(modname) - 1, modname); buf =read_text_file (filelist); -if (!buf) -/* not a module or .mod file missing - ignore */ +if (!buf) { +warn ("failed to open %s. cannot calculate checksum\n" , + filelist);return ; + } pos = buf; firstline =get_line (&pos);
Signature · 1 line
2.25.1
PATCH 6/29 modpost: avoid false-positive file open error
In reply to Masahiro YamadaOne problem of grab_file() is that it cannot distinguish the following two cases:
- I cannot read the file (the file does not exist, or read permission is not set)
- It can read the file, but the file size is zero
This is because grab_file() calls mmap(), which requires the mapped length is greater than 0. Hence, grab_file() fails for both cases.
If an empty header file were included from a module that requires checksum calculation, the following warning would be printed:
WARNING: modpost: could not open ...: Invalid argument
An empty file is a valid source file, so it should not fail.
Use read_text_file() instead. It can read a zero-length file. Then, parse_file() will succeed with doing nothing.
---
scripts/mod/sumversion.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
scripts/mod/sumversion.c +4 -2
@@ -258,10 +258,12 @@ static int parse_file(const char *fname, struct md4_ctx *md)char *file;unsigned long i, len; - file =grab_file (fname, &len); + file =read_text_file (fname);if (!file)return 0 ; + len =strlen (file); +for (i =0 ; i < len; i++) {/* Collapse and ignore \ and CR. */ if (file[i] =='\\' && (i+1 < len) && file[i+1 ] =='\n' ) {
@@ -287,7 +289,7 @@ static int parse_file(const char *fname, struct md4_ctx *md)add_char (file[i], md); } -release_file (file, len); +free (file);return 1 ; }/* Check whether the file is a static library or not */
Signature · 1 line
2.25.1
PATCH 7/29 modpost: use read_text_file() and get_line() for reading text files
In reply to Masahiro Yamadagrab_file() mmaps a file, but it is not so efficient here because get_next_line() copies every line to the temporary buffer anyway.
read_text_file() and get_line() are simpler. get_line() exploits the library function strchr().
---
scripts/mod/modpost.c | 15 ++++++++------- scripts/mod/sumversion.c | 11 ++++++----- 2 files changed, 14 insertions(+), 12 deletions(-)
scripts/mod/modpost.c +8 -7
@@ -2466,15 +2466,16 @@ static void write_if_changed(struct buffer *b, const char *fname)**/ static void read_dump (const char *fname,unsigned int kernel) { -unsigned long size, pos =0 ; -void *file =grab_file (fname, &size); -char *line; +char *buf, *pos, *line; -if (!file) + buf =read_text_file (fname); +if (!buf)/* No symbol versions, silently ignore */ return ; -while ((line =get_next_line (&pos, file, size))) { + pos = buf; + +while ((line =get_line (&pos))) {char *symname, *namespace, *modname, *d, *export;unsigned int crc;struct module *mod;
@@ -2509,10 +2510,10 @@ static void read_dump(const char *fname, unsigned int kernel)sym_set_crc (symname, crc);sym_update_namespace (symname, namespace); } -release_file (file, size); +free (buf);return ; fail: -release_file (file, size); +free (buf);fatal ("parse error in symbol dump file\n" ); }
scripts/mod/sumversion.c +6 -5
@@ -306,9 +306,8 @@ static int is_static_library(const char *objfile)* to figure out source files. */ static int parse_source_files (const char *objfile,struct md4_ctx *md) { -char *cmd, *file, *line, *dir; +char *cmd, *file, *line, *dir, *pos;const char *base; -unsigned long flen, pos =0 ;int dirlen, ret =0 , check_files =0 ; cmd =NOFAIL (malloc (strlen (objfile) +sizeof ("..cmd" )));
@@ -326,14 +325,16 @@ static int parse_source_files(const char *objfile, struct md4_ctx *md)strncpy (dir, objfile, dirlen); dir[dirlen] ='\0' ; - file =grab_file (cmd, &flen); + file =read_text_file (cmd);if (!file) {warn ("could not find %s for %s\n" , cmd, objfile); goto out; } + pos = file; +/* Sum all files in the same dir or subdirs. */ -while ((line =get_next_line (&pos, file, flen)) !=NULL ) { +while ((line =get_line (&pos))) {char * p = line;if (strncmp (line,"source_" ,sizeof ("source_" )-1 ) ==0 ) {
@@ -384,7 +385,7 @@ static int parse_source_files(const char *objfile, struct md4_ctx *md)/* Everyone parsed OK */ ret =1 ; out_file: -release_file (file, flen); +free (file); out:free (dir);free (cmd);
Signature · 1 line
2.25.1
PATCH 8/29 modpost: remove get_next_text() and make {grab,release_}file static
In reply to Masahiro Yamadaget_next_line() is no longer used. Remove.
grab_file() and release_file() are only used in modpost.c. Make them static.
---
scripts/mod/modpost.c | 38 ++------------------------------------ scripts/mod/modpost.h | 3 --- 2 files changed, 2 insertions(+), 39 deletions(-)
scripts/mod/modpost.c +2 -36
@@ -452,7 +452,7 @@ static void sym_set_crc(const char *name, unsigned int crc) s->crc_valid =1 ; } -void *grab_file (const char *filename,unsigned long *size) +static void *grab_file (const char *filename,unsigned long *size) {struct stat st;void *map =MAP_FAILED ;
@@ -474,41 +474,7 @@ void *grab_file(const char *filename, unsigned long *size)return map; } -/** -* Return a copy of the next line in a mmap'ed file. -* spaces in the beginning of the line is trimmed away. -* Return a pointer to a static buffer. -**/ -char *get_next_line (unsigned long *pos,void *file,unsigned long size) -{ -static char line[4096 ]; -int skip =1 ; -size_t len =0 ; -signed char *p = (signed char *)file + *pos; -char *s = line; - -for (; *pos < size ; (*pos)++) { -if (skip &&isspace (*p)) { - p++; -continue ; - } - skip =0 ; -if (*p !='\n' && (*pos < size)) { - len++; - *s++ = *p++; -if (len >4095 ) -break ;/* Too long, stop */ - }else { -/* End of string */ - *s ='\0' ; -return line; - } - } -/* End of buffer */ -return NULL ; -} - -void release_file (void *file,unsigned long size) +static void release_file (void *file,unsigned long size) {munmap (file, size); }
scripts/mod/modpost.h +0 -3
@@ -192,9 +192,6 @@ void get_src_version(const char *modname, char sum[], unsigned sumlen);/* from modpost.c */ char *read_text_file (const char *filename);char *get_line (char **stringp); -void *grab_file (const char *filename,unsigned long *size); -char *get_next_line (unsigned long *pos,void *file,unsigned long size); -void release_file (void *file,unsigned long size);enum loglevel {LOG_WARN ,
Signature · 1 line
2.25.1
PATCH 9/29 kbuild: disallow multi-word in M= or KBUILD_EXTMOD
In reply to Masahiro Yamada$(firstword ...) in scripts/Makefile.modpost was added by commit 3f3fd3c05585 ("[PATCH] kbuild: allow multi-word $M in Makefile.modpost") to build multiple external module directories.
This feature has been broken for a while. Remove the bitrotten code, and stop parsing if M or KBUILD_EXTMOD contains multiple words.
---
Makefile | 3 +++ scripts/Makefile.modpost | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-)
Makefile +3 -0
@@ -218,6 +218,9 @@ ifeq ("$(origin M)", "command line") KBUILD_EXTMOD := $(M) endif +$(if $(word 2, $(KBUILD_EXTMOD)), \ + $(error building multiple external modules is not supported)) + export KBUILD_CHECKSRC KBUILD_EXTMOD extmod-prefix = $(if $(KBUILD_EXTMOD),$(KBUILD_EXTMOD)/)
scripts/Makefile.modpost +1 -1
@@ -44,7 +44,7 @@ include include/config/auto.conf include scripts/Kbuild.include kernelsymfile := $(objtree)/Module.symvers -modulesymfile := $(firstword $(KBUILD_EXTMOD))/Module.symvers +modulesymfile := $(KBUILD_EXTMOD)/Module.symvers MODPOST = scripts/mod/modpost \ $(if $(CONFIG_MODVERSIONS),-m) \
Signature · 1 line
2.25.1
RE: [PATCH 09/29] kbuild: disallow multi-word in M= or KBUILD_EXTMOD
From: Masahiro Yamada
Quoted 7 lines
Sent: 17 May 2020 10:49
$(firstword ...) inscripts/Makefile.modpostwas added by commit3f3fd3c05585("[PATCH] kbuild: allow multi-word $M in Makefile.modpost") to build multiple external module directories.This feature has been broken for a while. Remove the bitrotten code, and stop parsing if M or
KBUILD_EXTMODcontains multiple words.
ISTR that one of the kernel documentation files says that it is possible to build multiple modules together in order to avoid 'faffing' with exported symbol lists.
So the docs need updating to match.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
Re: [PATCH 09/29] kbuild: disallow multi-word in M= or KBUILD_EXTMOD
Quoted 7 lines · David Laight
From: Masahiro Yamada
Quoted 5 lines
Sent: 17 May 2020 10:49
$(firstword ...) inscripts/Makefile.modpostwas added by commit3f3fd3c05585("[PATCH] kbuild: allow multi-word $M in Makefile.modpost") to build multiple external module directories.
Quoted 8 lines
Quoted 2 lines
This feature has been broken for a while. Remove the bitrotten code, and stop parsing if M or
KBUILD_EXTMODcontains multiple words.ISTR that one of the kernel documentation files says that it is possible to build multiple modules together in order to avoid 'faffing' with exported symbol lists.
So the docs need updating to match.
Do you remember which doc mentions it?
Quoted 6 lines
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
Signature · 2 lines
Best Regards Masahiro Yamada
PATCH 10/29 modpost: move -T option close to the modpost command
In reply to Masahiro YamadaThe '-T -' option reads the file list from stdin.
It is clearer to put it close to the piped command.
---
scripts/Makefile.modpost | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
scripts/Makefile.modpost +2 -2
@@ -66,7 +66,7 @@ __modpost: else -MODPOST += $(subst -i,-n,$(filter -i,$(MAKEFLAGS))) -s -T - \ +MODPOST += $(subst -i,-n,$(filter -i,$(MAKEFLAGS))) -s \ $(if $(KBUILD_NSDEPS),-d $(MODULES_NSDEPS)) ifeq ($(KBUILD_EXTMOD),)
@@ -88,7 +88,7 @@ modules := $(sort $(shell cat $(MODORDER))) # Read out modules.order instead of expanding $(modules) to pass in modpost. # Otherwise, allmodconfig would fail with "Argument list too long". quiet_cmd_modpost = MODPOST $(words $(modules)) modules - cmd_modpost = sed 's/ko$$/o/' $(MODORDER) | $(MODPOST) + cmd_modpost = sed 's/ko$$/o/' $(MODORDER) | $(MODPOST) -T - __modpost: $(call cmd,modpost)
Signature · 1 line
2.25.1
PATCH 11/29 modpost: pass -N option only for modules modpost
In reply to Masahiro YamadaThe built-in only code is not required to have MODULE_IMPORT_NS() to use symbols. So, the namespace is not checked for vmlinux(.o).
Do not pass the meaningless -N option to the first pass of modpost.
---
scripts/Makefile.modpost | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
scripts/Makefile.modpost +2 -1
@@ -53,7 +53,6 @@ MODPOST = scripts/mod/modpost \ $(if $(KBUILD_EXTMOD),$(addprefix -e ,$(KBUILD_EXTRA_SYMBOLS))) \ $(if $(KBUILD_EXTMOD),-o $(modulesymfile)) \ $(if $(CONFIG_SECTION_MISMATCH_WARN_ONLY),,-E) \ - $(if $(CONFIG_MODULE_ALLOW_MISSING_NAMESPACE_IMPORTS)$(KBUILD_NSDEPS),-N) \ $(if $(KBUILD_MODPOST_WARN),-w) ifdef MODPOST_VMLINUX
@@ -66,7 +65,9 @@ __modpost: else +# modpost options for modules (both in-kernel and external) MODPOST += $(subst -i,-n,$(filter -i,$(MAKEFLAGS))) -s \ + $(if $(CONFIG_MODULE_ALLOW_MISSING_NAMESPACE_IMPORTS)$(KBUILD_NSDEPS),-N) \ $(if $(KBUILD_NSDEPS),-d $(MODULES_NSDEPS)) ifeq ($(KBUILD_EXTMOD),)
Signature · 1 line
2.25.1
PATCH 12/29 modpost: move external module options
In reply to Masahiro YamadaSeparate out the external-only code for readability.
---
scripts/Makefile.modpost | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-)
scripts/Makefile.modpost +8 -4
@@ -44,14 +44,11 @@ include include/config/auto.conf include scripts/Kbuild.include kernelsymfile := $(objtree)/Module.symvers -modulesymfile := $(KBUILD_EXTMOD)/Module.symvers MODPOST = scripts/mod/modpost \ $(if $(CONFIG_MODVERSIONS),-m) \ $(if $(CONFIG_MODULE_SRCVERSION_ALL),-a) \ - $(if $(KBUILD_EXTMOD),-i,-o) $(kernelsymfile) \ - $(if $(KBUILD_EXTMOD),$(addprefix -e ,$(KBUILD_EXTRA_SYMBOLS))) \ - $(if $(KBUILD_EXTMOD),-o $(modulesymfile)) \ + $(if $(KBUILD_EXTMOD),,-o $(kernelsymfile)) \ $(if $(CONFIG_SECTION_MISMATCH_WARN_ONLY),,-E) \ $(if $(KBUILD_MODPOST_WARN),-w)
@@ -81,6 +78,13 @@ src := $(obj) # Include the module's Makefile to find KBUILD_EXTRA_SYMBOLS include $(if $(wildcard $(KBUILD_EXTMOD)/Kbuild), \ $(KBUILD_EXTMOD)/Kbuild, $(KBUILD_EXTMOD)/Makefile) + +# modpost options for external modules +MODPOST += \ + -i Module.symvers \ + $(addprefix -e ,$(KBUILD_EXTRA_SYMBOLS)) \ + -o $(KBUILD_EXTMOD)/Module.symvers + endif # find all modules listed in modules.order
Signature · 1 line
2.25.1
PATCH 13/29 modpost: load KBUILD_EXTRA_SYMBOLS files in order
In reply to Masahiro YamadaCurrently, modpost reads extra symbol dump files in the reverse order. If '-e foo -e bar' is given, modpost reads bar, foo, in this order.
This is probably not a big deal, but there is no good reason to reverse the order. Read files in the given order.
---
scripts/mod/modpost.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-)
scripts/mod/modpost.c +8 -8
@@ -2559,8 +2559,8 @@ int main(int argc, char **argv)int opt;int err;int n; -struct ext_sym_list *extsym_iter;struct ext_sym_list *extsym_start =NULL ; +struct ext_sym_list **extsym_iter = &extsym_start;while ((opt =getopt (argc, argv,"i:e:mnsT:o:awENd:" )) !=-1 ) {switch (opt) {
@@ -2570,11 +2570,9 @@ int main(int argc, char **argv)break ;case 'e' : external_module =1 ; - extsym_iter = -NOFAIL (malloc (sizeof (*extsym_iter))); - extsym_iter->next = extsym_start; - extsym_iter->file = optarg; - extsym_start = extsym_iter; + *extsym_iter =NOFAIL (calloc (1 ,sizeof (**extsym_iter))); + (*extsym_iter)->file = optarg; + extsym_iter = &(*extsym_iter)->next;break ;case 'm' : modversions =1 ;
@@ -2614,10 +2612,12 @@ int main(int argc, char **argv)if (kernel_read)read_dump (kernel_read,1 );while (extsym_start) { +struct ext_sym_list *tmp; +read_dump (extsym_start->file,0 ); - extsym_iter = extsym_start->next; + tmp = extsym_start->next;free (extsym_start); - extsym_start = extsym_iter; + extsym_start = tmp; }while (optind < argc)
Signature · 1 line
2.25.1
PATCH 14/29 modpost: track if the symbol origin is a dump file or ELF object
In reply to Masahiro YamadaThe meaning of sym->kernel is obscure; it is set for in-kernel symbols loaded from Modules.symver. This happens only when we are building external modules, and it is used to determine whether to dump symbols to $(KBUILD_EXTMOD)/Modules.symver
It is clearer to remember whether the symbol or module came from a dump file or ELF object.
This changes the KBUILD_EXTRA_SYMBOLS behavior. Previously, symbols loaded from KBUILD_EXTRA_SYMBOLS are accumulated into the current $(KBUILD_EXTMOD)/Modules.symver
Going forward, they will be only used to check symbol references, but not dumped into the current $(KBUILD_EXTMOD)/Modules.symver. I believe this makes more sense.
sym->vmlinux will have no user. Remove it too.
---
scripts/mod/modpost.c | 15 +++++---------- scripts/mod/modpost.h | 1 + 2 files changed, 6 insertions(+), 10 deletions(-)
scripts/mod/modpost.c +5 -10
@@ -197,9 +197,6 @@ struct symbol {int crc_valid;char *namespace; unsigned int weak:1; - unsigned int vmlinux:1; /* 1 if symbol is defined in vmlinux */ - unsigned int kernel:1; /* 1 if symbol is from kernel -* (only for external modules) **/ unsigned int is_static:1; /* 1 if symbol is not global */enum export export;/* Type of export */ char name[];
@@ -431,8 +428,6 @@ static struct symbol *sym_add_exported(const char *name, struct module *mod, } s->module = mod; - s->vmlinux =is_vmlinux (mod->name); - s->kernel =0 ; s->export = export;return s; }
@@ -2430,7 +2425,7 @@ static void write_if_changed(struct buffer *b, const char *fname)/* parse Module.symvers file. line format: * 0x12345678<tab>symbol<tab>module<tab>export<tab>namespace **/ -static void read_dump (const char *fname,unsigned int kernel) +static void read_dump (const char *fname) {char *buf, *pos, *line;
@@ -2469,9 +2464,9 @@ static void read_dump(const char *fname, unsigned int kernel) have_vmlinux =1 ; mod =new_module (modname); mod->skip =1 ; + mod->from_dump =1 ; } s =sym_add_exported (symname, mod,export_no (export)); - s->kernel = kernel; s->is_static =0 ;sym_set_crc (symname, crc);sym_update_namespace (symname, namespace);
@@ -2491,7 +2486,7 @@ static int dump_sym(struct symbol *sym) {if (!external_module)return 1 ; -if (sym->vmlinux || sym->kernel) +if (sym->module->from_dump)return 0 ;return 1 ; }
@@ -2610,11 +2605,11 @@ int main(int argc, char **argv) }if (kernel_read) -read_dump (kernel_read,1 ); +read_dump (kernel_read);while (extsym_start) {struct ext_sym_list *tmp; -read_dump (extsym_start->file,0 ); +read_dump (extsym_start->file); tmp = extsym_start->next;free (extsym_start); extsym_start = tmp;
scripts/mod/modpost.h +1 -0
@@ -119,6 +119,7 @@ struct module {const char *name;int gpl_compatible;struct symbol *unres; +int from_dump;/* 1 if module was loaded from *.symver */ int seen;int skip;int has_init;
Signature · 1 line
2.25.1
PATCH 15/29 modpost: allow to pass -i option multiple times remove -e option
In reply to Masahiro YamadaNow that there is no difference in the functionality of -i and -e, they can be unified.
Make modpost accept the -i option multiple times, then remove -e.
I will reuse -e for a different purpose.
---
scripts/Makefile.modpost | 2 +- scripts/mod/modpost.c | 9 +-------- 2 files changed, 2 insertions(+), 9 deletions(-)
scripts/Makefile.modpost +1 -1
@@ -82,7 +82,7 @@ include $(if $(wildcard $(KBUILD_EXTMOD)/Kbuild), \ # modpost options for external modules MODPOST += \ -i Module.symvers \ - $(addprefix -e ,$(KBUILD_EXTRA_SYMBOLS)) \ + $(addprefix -i ,$(KBUILD_EXTRA_SYMBOLS)) \ -o $(KBUILD_EXTMOD)/Module.symvers endif
scripts/mod/modpost.c +1 -8
@@ -2548,7 +2548,6 @@ int main(int argc, char **argv) {struct module *mod;struct buffer buf = { }; -char *kernel_read =NULL ;char *missing_namespace_deps =NULL ;char *dump_write =NULL , *files_source =NULL ;int opt;
@@ -2557,13 +2556,9 @@ int main(int argc, char **argv)struct ext_sym_list *extsym_start =NULL ;struct ext_sym_list **extsym_iter = &extsym_start; -while ((opt =getopt (argc, argv,"i:e:mnsT:o:awENd:" )) !=-1 ) { +while ((opt =getopt (argc, argv,"i:mnsT:o:awENd:" )) !=-1 ) {switch (opt) {case 'i' : - kernel_read = optarg; - external_module =1 ; -break ; -case 'e' : external_module =1 ; *extsym_iter =NOFAIL (calloc (1 ,sizeof (**extsym_iter))); (*extsym_iter)->file = optarg;
@@ -2604,8 +2599,6 @@ int main(int argc, char **argv) } } -if (kernel_read) -read_dump (kernel_read);while (extsym_start) {struct ext_sym_list *tmp;
Signature · 1 line
2.25.1
PATCH 16/29 modpost: rename ext_sym_list to dump_list
In reply to Masahiro YamadaThe -i option is used to include Modules.symver as well as files from $(KBUILD_EXTRA_SYMBOLS).
Make the struct and variable names more generic.
---
scripts/mod/modpost.c | 27 ++++++++++++++------------- 1 file changed, 14 insertions(+), 13 deletions(-)
scripts/mod/modpost.c +14 -13
@@ -2539,8 +2539,8 @@ static void write_namespace_deps_files(const char *fname)free (ns_deps_buf.p); } -struct ext_sym_list { -struct ext_sym_list *next; +struct dump_list { +struct dump_list *next;const char *file; };
@@ -2553,16 +2553,17 @@ int main(int argc, char **argv)int opt;int err;int n; -struct ext_sym_list *extsym_start =NULL ; -struct ext_sym_list **extsym_iter = &extsym_start; +struct dump_list *dump_read_start =NULL ; +struct dump_list **dump_read_iter = &dump_read_start;while ((opt =getopt (argc, argv,"i:mnsT:o:awENd:" )) !=-1 ) {switch (opt) {case 'i' : external_module =1 ; - *extsym_iter =NOFAIL (calloc (1 ,sizeof (**extsym_iter))); - (*extsym_iter)->file = optarg; - extsym_iter = &(*extsym_iter)->next; + *dump_read_iter = +NOFAIL (calloc (1 ,sizeof (**dump_read_iter))); + (*dump_read_iter)->file = optarg; + dump_read_iter = &(*dump_read_iter)->next;break ;case 'm' : modversions =1 ;
@@ -2599,13 +2600,13 @@ int main(int argc, char **argv) } } -while (extsym_start) { -struct ext_sym_list *tmp; +while (dump_read_start) { +struct dump_list *tmp; -read_dump (extsym_start->file); - tmp = extsym_start->next; -free (extsym_start); - extsym_start = tmp; +read_dump (dump_read_start->file); + tmp = dump_read_start->next; +free (dump_read_start); + dump_read_start = tmp; }while (optind < argc)
Signature · 1 line
2.25.1
PATCH 17/29 modpost: re-add -e to set external_module flag
In reply to Masahiro YamadaPreviously, the -i option had two functions; load a symbol dump file, and set the external_module flag.
I want to assign a dedicate option for each of them.
Going forward, the -i is used to load a symbol dump file, and the -e to set the external_module flag.
With this, we will be able to use -i for loading in-kernel symbols.
---
scripts/Makefile.modpost | 1 + scripts/mod/modpost.c | 6 ++++-- 2 files changed, 5 insertions(+), 2 deletions(-)
scripts/Makefile.modpost +1 -0
@@ -81,6 +81,7 @@ include $(if $(wildcard $(KBUILD_EXTMOD)/Kbuild), \ # modpost options for external modules MODPOST += \ + -e \ -i Module.symvers \ $(addprefix -i ,$(KBUILD_EXTRA_SYMBOLS)) \ -o $(KBUILD_EXTMOD)/Module.symvers
scripts/mod/modpost.c +4 -2
@@ -2556,10 +2556,12 @@ int main(int argc, char **argv)struct dump_list *dump_read_start =NULL ;struct dump_list **dump_read_iter = &dump_read_start; -while ((opt =getopt (argc, argv,"i:mnsT:o:awENd:" )) !=-1 ) { +while ((opt =getopt (argc, argv,"ei:mnsT:o:awENd:" )) !=-1 ) {switch (opt) { -case 'i' : +case 'e' : external_module =1 ; +break ; +case 'i' : *dump_read_iter =NOFAIL (calloc (1 ,sizeof (**dump_read_iter))); (*dump_read_iter)->file = optarg;
Signature · 1 line
2.25.1
PATCH 18/29 modpost: show warning if vmlinux is not found when processing modules
In reply to Masahiro Yamadacheck_exports() does not print warnings about unresolved symbols if vmlinux is missing because there would be too many.
This situation happens when you do 'make modules' from the clean tree, or compile external modules against a kernel tree that has not been completely built.
Not checking unresolved symbols is dangerous because you might be building useless modules. At least it should be noted.
---
scripts/mod/modpost.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
scripts/mod/modpost.c +7 -2
@@ -2004,8 +2004,6 @@ static void read_symbols(const char *modname) mod =new_module (modname); -/* When there's no vmlinux, don't print warnings about -* unresolved symbols (since there'll be too many ;) */ if (is_vmlinux (modname)) { have_vmlinux =1 ; mod->skip =1 ;
@@ -2617,6 +2615,13 @@ int main(int argc, char **argv)if (files_source)read_symbols_from_files (files_source); +/* +* When there's no vmlinux, don't print warnings about +* unresolved symbols (since there'll be too many ;) +*/ +if (!have_vmlinux) +warn ("Symbol info of vmlinux is missing. Unresolved symbol check will be entirely skipped.\n" ); + err =0 ;for (mod = modules; mod; mod = mod->next) {
Signature · 1 line
2.25.1
PATCH 19/29 modpost: show warning if it fails to read symbol dump file
In reply to Masahiro YamadaIf modpost fails to load a symbol dump file, it cannot check unresolved symbols, hence module dependency will not be added. Nor CRCs can be added.
Currently, external module builds check only $(objtree)/Module.symvers, but it should check files specified by KBUILD_EXTRA_SYMBOLS as well.
Print the warnings in modpost. The warning in Makefile is unneeded.
---
Makefile | 10 +--------- scripts/mod/modpost.c | 11 +++++++++-- 2 files changed, 10 insertions(+), 11 deletions(-)
Makefile +1 -9
@@ -1638,17 +1638,9 @@ else # KBUILD_EXTMOD # We are always building modules KBUILD_MODULES := 1 -PHONY += $(objtree)/Module.symvers -$(objtree)/Module.symvers: - @test -e $(objtree)/Module.symvers || ( \ - echo; \ - echo " WARNING: Symbol version dump $(objtree)/Module.symvers"; \ - echo " is missing; modules will have no dependencies and modversions."; \ - echo ) - build-dirs := $(KBUILD_EXTMOD) PHONY += modules -modules: descend $(objtree)/Module.symvers +modules: descend $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modpost PHONY += modules_install
scripts/mod/modpost.c +9 -2
@@ -41,6 +41,8 @@ static int sec_mismatch_fatal = 0;static int ignore_missing_files;/* If set to 1, only warn (instead of error) about missing ns imports */ static int allow_missing_ns_imports; +/* Set when at list one dump file is missing */ +static int missing_dump_file;enum export { export_plain, export_unused, export_gpl,
@@ -2428,9 +2430,11 @@ static void read_dump(const char *fname)char *buf, *pos, *line; buf =read_text_file (fname); -if (!buf) -/* No symbol versions, silently ignore */ +if (!buf) { +warn ("failed to read '%s'\n" , fname); + missing_dump_file =1 ;return ; + } pos = buf;
@@ -2615,6 +2619,9 @@ int main(int argc, char **argv)if (files_source)read_symbols_from_files (files_source); +if (missing_dump_file) +warn ("Symbol dump file is missing. Modules may not have dependencies or movversions.\n" ); +/* * When there's no vmlinux, don't print warnings about * unresolved symbols (since there'll be too many ;)
Signature · 1 line
2.25.1
PATCH 20/29 modpost: generate vmlinux.symvers and reuse it for the second modpost
In reply to Masahiro YamadaThe full build runs modpost twice, first for vmlinux.o and second for modules.
The first pass dumps all the vmlinux symbols into Module.symvers, but the second pass parses vmlinux again instead of reusing the dump file, presumably because it needs to avoid accumulating stale symbols.
Loading symbol info from a dump file is faster than parsing an ELF object. Besides, modpost deals with various issues to parse vmlinux in the second pass.
A solution is to make the first pass dumps symbols into a separate file, vmlinux.symvers. The second pass reads it, and parses module .o files. The merged symbol information is dumped into Module.symvers as before.
This makes further modpost cleanups possible.
Also, it fixes the problem of 'make vmlinux', which previously overwrote Module.symvers, throwing away module symbols.
I slightly touched scripts/link-vmlinux.sh so that vmlinux is re-linked when you cross this commit. Otherwise, vmlinux.symvers would not be generated.
---
.gitignore | 1 + Makefile | 2 +- scripts/Makefile.modpost | 11 +++++++---- scripts/link-vmlinux.sh | 2 -- 4 files changed, 9 insertions(+), 7 deletions(-)
.gitignore +1 -0
@@ -56,6 +56,7 @@ modules.order /linux /vmlinux /vmlinux.32 +/vmlinux.symvers /vmlinux-gdb.py /vmlinuz /System.map
Makefile +1 -1
@@ -1405,7 +1405,7 @@ endif # CONFIG_MODULES # make distclean Remove editor backup files, patch leftover files and the like # Directories & files removed with 'make clean' -CLEAN_FILES += include/ksym \ +CLEAN_FILES += include/ksym vmlinux.symvers \ modules.builtin modules.builtin.modinfo modules.nsdeps # Directories & files removed with 'make mrproper'
scripts/Makefile.modpost +7 -4
@@ -43,17 +43,17 @@ __modpost: include include/config/auto.conf include scripts/Kbuild.include -kernelsymfile := $(objtree)/Module.symvers - MODPOST = scripts/mod/modpost \ $(if $(CONFIG_MODVERSIONS),-m) \ $(if $(CONFIG_MODULE_SRCVERSION_ALL),-a) \ - $(if $(KBUILD_EXTMOD),,-o $(kernelsymfile)) \ $(if $(CONFIG_SECTION_MISMATCH_WARN_ONLY),,-E) \ $(if $(KBUILD_MODPOST_WARN),-w) ifdef MODPOST_VMLINUX +# modpost options for vmlinux.o +MODPOST += -o vmlinux.symvers + quiet_cmd_modpost = MODPOST vmlinux.o cmd_modpost = $(MODPOST) vmlinux.o
@@ -68,7 +68,10 @@ MODPOST += $(subst -i,-n,$(filter -i,$(MAKEFLAGS))) -s \ $(if $(KBUILD_NSDEPS),-d $(MODULES_NSDEPS)) ifeq ($(KBUILD_EXTMOD),) -MODPOST += $(wildcard vmlinux) + +# modpost options for in-kernel modules +MODPOST += -i vmlinux.symvers -o Module.symvers + else # set src + obj - they may be used in the modules's Makefile
scripts/link-vmlinux.sh +0 -2
@@ -218,8 +218,6 @@ on_signals() } trap on_signals HUP INT QUIT TERM -# -# # Use "make V=1" to debug this script case "${KBUILD_VERBOSE}" in *1*)
Signature · 1 line
2.25.1
PATCH 21/29 modpost: remove -s option
In reply to Masahiro YamadaThe -s option was added by commit 8d8d8289df65 ("kbuild: do not do section mismatch checks on vmlinux in 2nd pass").
Now that the second pass does not parse vmlinux, this option is unneeded.
---
scripts/Makefile.modpost | 2 +- scripts/mod/modpost.c | 10 ++-------- 2 files changed, 3 insertions(+), 9 deletions(-)
scripts/Makefile.modpost +1 -1
@@ -63,7 +63,7 @@ __modpost: else # modpost options for modules (both in-kernel and external) -MODPOST += $(subst -i,-n,$(filter -i,$(MAKEFLAGS))) -s \ +MODPOST += $(subst -i,-n,$(filter -i,$(MAKEFLAGS))) \ $(if $(CONFIG_MODULE_ALLOW_MISSING_NAMESPACE_IMPORTS)$(KBUILD_NSDEPS),-N) \ $(if $(KBUILD_NSDEPS),-d $(MODULES_NSDEPS))
scripts/mod/modpost.c +2 -8
@@ -30,8 +30,6 @@ static int have_vmlinux = 0;static int all_versions =0 ;/* If we are modposting external module set to 1 */ static int external_module =0 ; -/* Warn about section mismatch in vmlinux if set to 1 */ -static int vmlinux_section_warnings =1 ;/* Only warn about unresolved symbols */ static int warn_unresolved =0 ;/* How a symbol is exported */
@@ -2068,8 +2066,7 @@ static void read_symbols(const char *modname) } } -if (!is_vmlinux (modname) || vmlinux_section_warnings) -check_sec_ref (mod, modname, &info); +check_sec_ref (mod, modname, &info);if (!is_vmlinux (modname)) { version =get_modinfo (&info,"version" );
@@ -2558,7 +2555,7 @@ int main(int argc, char **argv)struct dump_list *dump_read_start =NULL ;struct dump_list **dump_read_iter = &dump_read_start; -while ((opt =getopt (argc, argv,"ei:mnsT:o:awENd:" )) !=-1 ) { +while ((opt =getopt (argc, argv,"ei:mnT:o:awENd:" )) !=-1 ) {switch (opt) {case 'e' : external_module =1 ;
@@ -2581,9 +2578,6 @@ int main(int argc, char **argv)case 'a' : all_versions =1 ;break ; -case 's' : - vmlinux_section_warnings =0 ; -break ;case 'T' : files_source = optarg;break ;
Signature · 1 line
2.25.1
PATCH 22/29 modpost: remove mod->is_dot_o struct member
In reply to Masahiro YamadaPreviously, there were two cases where mod->is_dot_o is unset:
[1] the executable 'vmlinux' in the second pass of modpost
[2] modules loaded by read_dump()
I think [1] was intended usage to distinguish 'vmlinux.o' and 'vmlinux'. Now that modpost does not parse the executable 'vmlinux', this case does not happen.
[2] is obscure, maybe a bug. Module.symver stores module paths without extension. So, none of modules loaded by read_dump() has the .o suffix, and new_module() unsets ->is_dot_o. Anyway, it is not a big deal because handle_symbol() is not called for the case.
To sum up, all the parsed ELF files are .o files.
mod->is_dot_o is unneeded.
---
scripts/mod/modpost.c | 14 ++------------ scripts/mod/modpost.h | 1 - 2 files changed, 2 insertions(+), 13 deletions(-)
scripts/mod/modpost.c +2 -12
@@ -171,10 +171,8 @@ static struct module *new_module(const char *modname) p =NOFAIL (strdup (modname));/* strip trailing .o */ -if (strends (p,".o" )) { +if (strends (p,".o" )) p[strlen (p) -2 ] ='\0' ; - mod->is_dot_o =1 ; - }/* add to list */ mod->name = p;
@@ -702,8 +700,7 @@ static void handle_symbol(struct module *mod, struct elf_info *info,enum export export;const char *name; -if ((!is_vmlinux (mod->name) || mod->is_dot_o) && -strstarts (symname,"__ksymtab" )) +if (strstarts (symname,"__ksymtab" )) export =export_from_secname (info,get_secindex (info, sym));else export =export_from_sec (info,get_secindex (info, sym));
@@ -2661,13 +2658,6 @@ int main(int argc, char **argv)struct symbol *s;for (s = symbolhash[n]; s; s = s->next) { -/* -* Do not check "vmlinux". This avoids the same warnings -* shown twice, and false-positives for ARCH=um. -*/ -if (is_vmlinux (s->module->name) && !s->module->is_dot_o) -continue ; -if (s->is_static) warn("\"%s\" [%s] is a static %s\n", s->name, s->module->name,
scripts/mod/modpost.h +0 -1
@@ -126,7 +126,6 @@ struct module {int has_cleanup;struct buffer dev_table_buf;char srcversion[25 ]; -int is_dot_o;// Missing namespace dependencies struct namespace_list *missing_namespaces;// Actual imported namespaces
Signature · 1 line
2.25.1
PATCH 23/29 modpost: remove is_vmlinux() call in check_for_{gpl_usage,unused}()
In reply to Masahiro Yamadacheck_exports() is never called for vmlinux because mod->skip is set for vmlinux.
Hence, check_for_gpl_usage() and check_for_unused() are not called for vmlinux, either. is_vmlinux() is always false here.
Remove the is_vmlinux() calls, and hard-code the ".ko" suffix.
---
scripts/mod/modpost.c | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-)
scripts/mod/modpost.c +8 -12
@@ -2134,20 +2134,18 @@ void buf_write(struct buffer *buf, const char *s, int len)static void check_for_gpl_usage (enum export exp,const char *m,const char *s) { -const char *e =is_vmlinux (m) ?"" :".ko" ; -switch (exp) {case export_gpl: -fatal ("GPL-incompatible module %s%s " -"uses GPL-only symbol '%s'\n" , m, e, s); +fatal ("GPL-incompatible module %s.ko uses GPL-only symbol '%s'\n" , + m, s);break ;case export_unused_gpl: -fatal ("GPL-incompatible module %s%s " -"uses GPL-only symbol marked UNUSED '%s'\n" , m, e, s); +fatal ("GPL-incompatible module %s.ko uses GPL-only symbol marked UNUSED '%s'\n" , + m, s);break ;case export_gpl_future: -warn ("GPL-incompatible module %s%s " -"uses future GPL-only symbol '%s'\n" , m, e, s); +warn ("GPL-incompatible module %s.ko uses future GPL-only symbol '%s'\n" , + m, s);break ;case export_plain:case export_unused:
@@ -2159,13 +2157,11 @@ static void check_for_gpl_usage(enum export exp, const char *m, const char *s)static void check_for_unused (enum export exp,const char *m,const char *s) { -const char *e =is_vmlinux (m) ?"" :".ko" ; -switch (exp) {case export_unused:case export_unused_gpl: -warn ("module %s%s " -"uses symbol '%s' marked UNUSED\n" , m, e, s); +warn ("module %s.ko uses symbol '%s' marked UNUSED\n" , + m, s);break ;default :/* ignore */
Signature · 1 line
2.25.1
PATCH 24/29 modpost: add mod->is_vmlinux struct member
In reply to Masahiro Yamadais_vmlinux() is called in several places to check whether the current module is vmlinux or not.
It is faster and clearer to check mod->is_vmlinux flag.
---
scripts/mod/modpost.c | 19 ++++++++++--------- scripts/mod/modpost.h | 1 + 2 files changed, 11 insertions(+), 9 deletions(-)
scripts/mod/modpost.c +10 -9
@@ -176,6 +176,7 @@ static struct module *new_module(const char *modname)/* add to list */ mod->name = p; + mod->is_vmlinux =is_vmlinux (modname); mod->gpl_compatible =-1 ; mod->next = modules; modules = mod;
@@ -417,11 +418,11 @@ static struct symbol *sym_add_exported(const char *name, struct module *mod,if (!s) { s =new_symbol (name, mod, export); - }else if (!external_module ||is_vmlinux (s->module->name) || + }else if (!external_module || s->module->is_vmlinux || s->module == mod) {warn ("%s: '%s' exported twice. Previous export was in %s%s\n" , mod->name, name, s->module->name, -is_vmlinux (s->module->name) ?"" :".ko" ); + s->module->is_vmlinux ?"" :".ko" );return s; }
@@ -678,7 +679,7 @@ static void handle_modversion(const struct module *mod,if (sym->st_shndx ==SHN_UNDEF ) {warn ("EXPORT symbol \"%s\" [%s%s] version generation failed, symbol will not be versioned.\n" , - symname, mod->name,is_vmlinux (mod->name) ?"" :".ko" ); + symname, mod->name, mod->is_vmlinux ?"" :".ko" );return ; }
@@ -2001,12 +2002,12 @@ static void read_symbols(const char *modname) mod =new_module (modname); -if (is_vmlinux (modname)) { +if (mod->is_vmlinux) { have_vmlinux =1 ; mod->skip =1 ; } -if (!is_vmlinux (modname)) { +if (!mod->is_vmlinux) { license =get_modinfo (&info,"license" );if (!license)warn ("missing MODULE_LICENSE() in %s\n" , modname);
@@ -2065,7 +2066,7 @@ static void read_symbols(const char *modname)check_sec_ref (mod, modname, &info); -if (!is_vmlinux (modname)) { +if (!mod->is_vmlinux) { version =get_modinfo (&info,"version" );if (version || all_versions) get_src_version(modname, mod->srcversion,
@@ -2335,7 +2336,7 @@ static void add_depends(struct buffer *b, struct module *mod)/* Clear ->seen flag of modules that own symbols needed by this. */ for (s = mod->unres; s; s = s->next)if (s->module) - s->module->seen =is_vmlinux (s->module->name); + s->module->seen = s->module->is_vmlinux;buf_printf (b,"\n" );buf_printf (b,"MODULE_INFO(depends, \"" );
@@ -2452,9 +2453,9 @@ static void read_dump(const char *fname) goto fail; mod =find_module (modname);if (!mod) { -if (is_vmlinux (modname)) - have_vmlinux =1 ; mod =new_module (modname); +if (mod->is_vmlinux) + have_vmlinux =1 ; mod->skip =1 ; mod->from_dump =1 ; }
scripts/mod/modpost.h +1 -0
@@ -120,6 +120,7 @@ struct module {int gpl_compatible;struct symbol *unres;int from_dump;/* 1 if module was loaded from *.symver */ +int is_vmlinux;int seen;int skip;int has_init;
Signature · 1 line
2.25.1
PATCH 25/29 modpost: remove mod->skip struct member
In reply to Masahiro YamadaThe meaning of 'skip' is obscure since it does not explain "what to skip".
mod->skip is set when it is vmlinux or the module info came from a dump file.
So, mod->skip is equivalent to (mod->is_vmlinux || mod->from_dump).
For the check in write_namespace_deps_files(), mod->is_vmlinux is unneeded because the -d option is not passed in the first pass of modpost.
---
scripts/mod/modpost.c | 9 +++------ scripts/mod/modpost.h | 1 - 2 files changed, 3 insertions(+), 7 deletions(-)
scripts/mod/modpost.c +3 -6
@@ -2002,10 +2002,8 @@ static void read_symbols(const char *modname) mod =new_module (modname); -if (mod->is_vmlinux) { +if (mod->is_vmlinux) have_vmlinux =1 ; - mod->skip =1 ; - }if (!mod->is_vmlinux) { license =get_modinfo (&info,"license" );
@@ -2456,7 +2454,6 @@ static void read_dump(const char *fname) mod =new_module (modname);if (mod->is_vmlinux) have_vmlinux =1 ; - mod->skip =1 ; mod->from_dump =1 ; } s =sym_add_exported (symname, mod,export_no (export));
@@ -2517,7 +2514,7 @@ static void write_namespace_deps_files(const char *fname)for (mod = modules; mod; mod = mod->next) { -if (mod->skip || !mod->missing_namespaces) +if (mod->from_dump || !mod->missing_namespaces)continue ;buf_printf (&ns_deps_buf,"%s.ko:" , mod->name);
@@ -2622,7 +2619,7 @@ int main(int argc, char **argv)for (mod = modules; mod; mod = mod->next) {char fname[PATH_MAX ]; -if (mod->skip) +if (mod->is_vmlinux || mod->from_dump)continue ; buf.pos =0 ;
scripts/mod/modpost.h +0 -1
@@ -122,7 +122,6 @@ struct module {int from_dump;/* 1 if module was loaded from *.symver */ int is_vmlinux;int seen; -int skip;int has_init;int has_cleanup;struct buffer dev_table_buf;
Signature · 1 line
2.25.1
PATCH 26/29 modpost: set have_vmlinux in new_module()
In reply to Masahiro YamadaSet have_vmlinux flag in a single place.
---
scripts/mod/modpost.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-)
scripts/mod/modpost.c +3 -5
@@ -181,6 +181,9 @@ static struct module *new_module(const char *modname) mod->next = modules; modules = mod; +if (mod->is_vmlinux) + have_vmlinux =1 ; +return mod; }
@@ -2002,9 +2005,6 @@ static void read_symbols(const char *modname) mod =new_module (modname); -if (mod->is_vmlinux) - have_vmlinux =1 ; -if (!mod->is_vmlinux) { license =get_modinfo (&info,"license" );if (!license)
@@ -2452,8 +2452,6 @@ static void read_dump(const char *fname) mod =find_module (modname);if (!mod) { mod =new_module (modname); -if (mod->is_vmlinux) - have_vmlinux =1 ; mod->from_dump =1 ; } s =sym_add_exported (symname, mod,export_no (export));
Signature · 1 line
2.25.1
PATCH 27/29 modpost: strip .o from modname before calling new_module()
In reply to Masahiro Yamadanew_module() conditionally strips the .o because the modname has .o suffix when it is called from read_symbols(), but no .o when it is called from read_dump().
It is clearer to strip .o in read_symbols().
I also used flexible-array for mod->name.
---
scripts/mod/modpost.c | 20 +++++++++++--------- scripts/mod/modpost.h | 2 +- 2 files changed, 12 insertions(+), 10 deletions(-)
scripts/mod/modpost.c +11 -9
@@ -164,18 +164,12 @@ static struct module *find_module(const char *modname)static struct module *new_module (const char *modname) {struct module *mod; -char *p; - mod =NOFAIL (malloc (sizeof (*mod))); + mod =NOFAIL (malloc (sizeof (*mod) +strlen (modname) +1 ));memset (mod,0 ,sizeof (*mod)); - p =NOFAIL (strdup (modname)); - -/* strip trailing .o */ -if (strends (p,".o" )) - p[strlen (p) -2 ] ='\0' ;/* add to list */ - mod->name = p; +strcpy (mod->name, modname); mod->is_vmlinux =is_vmlinux (modname); mod->gpl_compatible =-1 ; mod->next = modules;
@@ -2003,7 +1997,15 @@ static void read_symbols(const char *modname)if (!parse_elf (&info, modname))return ; - mod =new_module (modname); + { +char *tmp; + +/* strip trailing .o */ + tmp =NOFAIL (strdup (modname)); + tmp[strlen (tmp) -2 ] ='\0' ; + mod =new_module (tmp); +free (tmp); + }if (!mod->is_vmlinux) { license =get_modinfo (&info,"license" );
scripts/mod/modpost.h +1 -1
@@ -116,7 +116,6 @@ struct namespace_list {struct module {struct module *next; -const char *name;int gpl_compatible;struct symbol *unres;int from_dump;/* 1 if module was loaded from *.symver */
@@ -130,6 +129,7 @@ struct module {struct namespace_list *missing_namespaces;// Actual imported namespaces struct namespace_list *imported_namespaces; +char name[]; };struct elf_info {
Signature · 1 line
2.25.1
PATCH 28/29 modpost: remove is_vmlinux() helper
In reply to Masahiro YamadaNow that is_vmlinux() is called only in new_module(), we can inline the function call.
modname is the basename with '.o' is stripped. No need to compare it with 'vmlinux.o'.
vmlinux is always located at the current working directory. No need to strip the directory path.
---
scripts/mod/modpost.c | 16 +--------------- 1 file changed, 1 insertion(+), 15 deletions(-)
scripts/mod/modpost.c +1 -15
@@ -90,20 +90,6 @@ static inline bool strends(const char *str, const char *postfix)return strcmp (str +strlen (str) -strlen (postfix), postfix) ==0 ; } -static int is_vmlinux (const char *modname) -{ -const char *myname; - - myname =strrchr (modname,'/' ); -if (myname) - myname++; -else - myname = modname; - -return (strcmp (myname,"vmlinux" ) ==0 ) || - (strcmp (myname,"vmlinux.o" ) ==0 ); -} -void *do_nofail (void *ptr,const char *expr) {if (!ptr)
@@ -170,7 +156,7 @@ static struct module *new_module(const char *modname)/* add to list */ strcpy (mod->name, modname); - mod->is_vmlinux =is_vmlinux (modname); + mod->is_vmlinux = (strcmp (modname,"vmlinux" ) ==0 ); mod->gpl_compatible =-1 ; mod->next = modules; modules = mod;
Signature · 1 line
2.25.1
PATCH 29/29 modpost: change elf_info->size to size_t
In reply to Masahiro YamadaAlign with the mmap / munmap APIs.
---
scripts/mod/modpost.c | 9 ++++----- scripts/mod/modpost.h | 2 +- 2 files changed, 5 insertions(+), 6 deletions(-)
scripts/mod/modpost.c +4 -5
@@ -429,7 +429,7 @@ static void sym_set_crc(const char *name, unsigned int crc) s->crc_valid =1 ; } -static void *grab_file (const char *filename,unsigned long *size) +static void *grab_file (const char *filename,size_t *size) {struct stat st;void *map =MAP_FAILED ;
@@ -451,7 +451,7 @@ static void *grab_file(const char *filename, unsigned long *size)return map; } -static void release_file (void *file,unsigned long size) +static void release_file (void *file,size_t size) {munmap (file, size); }
@@ -507,9 +507,8 @@ static int parse_elf(struct elf_info *info, const char *filename)/* Check if file offset is correct */ if (hdr->e_shoff > info->size) { -fatal ("section header offset=%lu in file '%s' is bigger than " -"filesize=%lu\n" , (unsigned long )hdr->e_shoff, - filename, info->size); +fatal ("section header offset=%lu in file '%s' is bigger than filesize=%zu\n" , + (unsigned long )hdr->e_shoff, filename, info->size);return 0 ; }
scripts/mod/modpost.h +1 -1
@@ -133,7 +133,7 @@ struct module { };struct elf_info { -unsigned long size; +size_t size;Elf_Ehdr *hdr;Elf_Shdr *sechdrs;Elf_Sym *symtab_start;
Signature · 1 line
2.25.1