lib/vsprintf: Make no_hash_pointers take effect early
Subsystems (2)
PATCH v3 1/1 lib/vsprintf: Make no_hash_pointers take effect early
From: Kaitao Cheng <[email protected]>
The no_hash_pointers boot parameter is now handled as an alias for hash_pointers=never. However, hash_pointers=never only records the selected mode during early parameter parsing, and no_hash_pointers is not updated until hash_pointers_finalize() runs later from SLUB init.
This leaves a window during very early boot where %p output is still hashed even though the user explicitly requested unhashed pointers with no_hash_pointers or hash_pointers=never.
Set no_hash_pointers as soon as the "never" mode is parsed. The later hash_pointers_finalize() call still keeps the final policy decision in one place, but explicit requests to disable pointer hashing now take effect for early boot users too.
---
Changes in v3 (Petr Mladek):
- Avoid stale no_hash_pointers state when hash_pointers overrides it.
Changes in v2 (Andy Shevchenko):
- Add a description of the kernel documentation to the commit log.
Link to v2:
https://lore.kernel.org/all/[email protected]/
Link to v1:
https://lore.kernel.org/all/[email protected]/
--- lib/vsprintf.c | 4 ++++ 1 file changed, 4 insertions(+)
lib/vsprintf.c +4 -0
@@ -2352,6 +2352,9 @@ void __init hash_pointers_finalize(bool slub_debug)static int __init hash_pointers_mode_parse (char *str) { +/* Avoid stale no_hash_pointers state when hash_pointers overrides it */ + no_hash_pointers = false; +if (!str) {pr_warn ("Hash pointers mode empty; falling back to auto.\n" ); hash_pointers_mode =HASH_PTR_AUTO ;
@@ -2361,6 +2364,7 @@ static int __init hash_pointers_mode_parse(char *str) }else if (strcmp (str,"never" ) ==0 ) {pr_info ("Hash pointers mode set to never.\n" ); hash_pointers_mode =HASH_PTR_NEVER ; + no_hash_pointers = true; }else if (strcmp (str,"always" ) ==0 ) {pr_info ("Hash pointers mode set to always.\n" ); hash_pointers_mode =HASH_PTR_ALWAYS ;
Signature · 1 line
2.50.1 (Apple Git-155)
Quoted 11 lines · Kaitao Cheng
From: Kaitao Cheng <[email protected]>
The
no_hash_pointersboot parameter is now handled as an alias for hash_pointers=never. However, hash_pointers=never only records the selected mode during early parameter parsing, andno_hash_pointersis not updated untilhash_pointers_finalize()runs later from SLUB init.This leaves a window during very early boot where %p output is still hashed even though the user explicitly requested unhashed pointers with
no_hash_pointersor hash_pointers=never.
Quoted 6 lines
Set
no_hash_pointersas soon as the "never" mode is parsed. The laterhash_pointers_finalize()call still keeps the final policy decision in one place, but explicit requests to disable pointer hashing now take effect for early boot users too.Signed-off-by Kaitao Cheng <[email protected]>
LGTM, feel free to use:
I am going to queue it via printk git tree for 7.3. I have already sent printk changes for 7.2 and this does not look serious enough to create yet another pull request. I could still add it to a pull request for 7.2 if a more serious fix appeared.
Best Regards,
Petr
Petr Mladek
maintainerQuoted 11 lines · Kaitao Cheng
From: Kaitao Cheng <[email protected]>
The
no_hash_pointersboot parameter is now handled as an alias for hash_pointers=never. However, hash_pointers=never only records the selected mode during early parameter parsing, andno_hash_pointersis not updated untilhash_pointers_finalize()runs later from SLUB init.This leaves a window during very early boot where %p output is still hashed even though the user explicitly requested unhashed pointers with
no_hash_pointersor hash_pointers=never.
Quoted 6 lines
Set
no_hash_pointersas soon as the "never" mode is parsed. The laterhash_pointers_finalize()call still keeps the final policy decision in one place, but explicit requests to disable pointer hashing now take effect for early boot users too.Signed-off-by Kaitao Cheng <[email protected]>
JFYI, the patch has been pushed into printk/linux.git, branch for-7.3-trivial.
As the branch name suggests, it is intended for 7.3. But as a trivial change, I could add it to a pull request for some 7.2-rcX when I need to create one because of a more important fix.
Best Regards,
Petr