linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Jim Cromie <jim.cromie@gmail.com>
To: Jason Baron <jbaron@akamai.com>, linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org, Jim Cromie <jim.cromie@gmail.com>
Subject: [RFC PATCH v5 27/28] dyndbg: unionize _ddebug*_headers with struct _ddebug*
Date: Tue, 11 May 2021 12:50:56 -0600	[thread overview]
Message-ID: <20210511185057.3815777-28-jim.cromie@gmail.com> (raw)
In-Reply-To: <20210511185057.3815777-1-jim.cromie@gmail.com>

Up until now, to create our header record (pair), we reused struct
_ddebug(|_sites), initializing the pair with "impossible" field values
(for a pr_debug() site), which makes it recognizable as a header not a
pr-debug callsite.

Now we want to specialize the header; differentiate it from its model,
then we can drop _ddebug.site, and use header.site indirectly.

This commit has several elements combining:

Update DEFINE_DYNAMIC_DEBUG_TABLE by replacing "struct _ddebug*" with
"union _ddebug*_header".  This change places the new object types into
our 2 elf sections.

our 2 new union types have 2 jobs:

a.fit properly in the section/array.  We do this by putting the struct
  inside the union, making the unions the same size as their contained
  structs.  This was sufficient to fix a data-misalignment crash on
  2nd loop (ie 1st record after header) in _init.

b.unions need the same field defns as their models, as refd in:
  DEFINE_DYNAMIC_DEBUG_TABLE() - inits
  is_dyndbg_header_pair(i,s)   - validates that init
  use anonymous struct.

RFC: My "anonymous union/struct fu" feels incomplete.  The struct in b
must contain all the fields of its outer union's own contained struct,
maybe even in the same order (tho not obvious yet).

Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
---
 include/linux/dynamic_debug.h | 31 ++++++++++++++++++++++++++++---
 1 file changed, 28 insertions(+), 3 deletions(-)

diff --git a/include/linux/dynamic_debug.h b/include/linux/dynamic_debug.h
index 46901b5348ee..9ca413985fb2 100644
--- a/include/linux/dynamic_debug.h
+++ b/include/linux/dynamic_debug.h
@@ -36,6 +36,16 @@ struct _ddebug_site {
 	const char *function;
 } __aligned(8);
 
+/* verbatim copy of struct _ddebug_site. nothing to save here. */
+union _ddebug_site_header {
+	struct _ddebug_site __;		/* inherit footprint */
+	struct {			/* header validation */
+		const char *modname;
+		const char *filename;
+		const char *function;
+	};
+};
+
 struct _ddebug {
 	SITE_CHK_(struct _ddebug_site *site;)
 	/* format is always needed, lineno shares word with flags */
@@ -74,6 +84,21 @@ struct _ddebug {
 #endif
 } __aligned(8);
 
+/*
+ * specialized version of struct _ddebug.  The only field we NEED is
+ * .site, since keeping it here allows dropping from struct _ddebug
+ * itself.  format, lineno are used to validate header records.
+ */
+union _ddebug_header {
+	struct _ddebug __;		/* inherit footprint */
+	struct {			/* header validation */
+		union _ddebug_site_header *site;
+		const char *format;
+		const unsigned lineno:18;
+		unsigned _index:14;
+		unsigned int flags:8;
+	};
+};
 
 #if defined(CONFIG_DYNAMIC_DEBUG_CORE)
 
@@ -136,10 +161,10 @@ void __dynamic_ibdev_dbg(struct _ddebug *descriptor,
  */
 #define DEFINE_DYNAMIC_DEBUG_TABLE()				\
 	/* forward decl, allowing loopback pointer */		\
-	__weak struct _ddebug __used __aligned(8)		\
+	__weak union _ddebug_header __used __aligned(8)		\
 		__section(".gnu.linkonce.dyndbg")		\
 		_LINKONCE_dyndbg_header;			\
-	__weak struct _ddebug_site __used __aligned(8)		\
+	__weak union _ddebug_site_header __used __aligned(8)	\
 		__section(".gnu.linkonce.dyndbg_site")		\
 	_LINKONCE_dyndbg_site_header = {			\
 		.modname = KBUILD_MODNAME,			\
@@ -147,7 +172,7 @@ void __dynamic_ibdev_dbg(struct _ddebug *descriptor,
 		/* forced pointer loopback, for distinction */	\
 		.filename = (void *) &_LINKONCE_dyndbg_header	\
 	};							\
-	__weak struct _ddebug __used __aligned(8)		\
+	__weak union _ddebug_header __used __aligned(8)		\
 		__section(".gnu.linkonce.dyndbg")		\
 	_LINKONCE_dyndbg_header = {				\
 		SITE_INIT_(_LINKONCE_dyndbg_site_header)	\
-- 
2.31.1



  parent reply	other threads:[~2021-05-11 18:52 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-11 18:50 [RFC PATCH v5 00/28] dynamic debug diet plan Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 01/28] dyndbg: avoid calling dyndbg_emit_prefix when it has no work Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 02/28] dyndbg: drop uninformative vpr_info Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 03/28] dyndbg: split struct _ddebug's display fields to new _ddebug_site Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 04/28] dyndbg: __init iterate over __dyndbg & __dyndbg_site in parallel Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 05/28] dyndbg: refactor part of ddebug_change to ddebug_match_site Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 06/28] dyndbg: accept null site in ddebug_match_site Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 07/28] dyndbg: hoist ->site out of ddebug_match_site Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 08/28] dyndbg: accept null site in ddebug_change Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 09/28] dyndbg: accept null site in dynamic_emit_prefix Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 10/28] dyndbg: accept null site in ddebug_proc_show Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 11/28] dyndbg: refactor ddebug_alter_site out of ddebug_change Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 12/28] dyndbg: allow deleting site info via control interface Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 13/28] dyndbg+module: expose ddebug_sites to modules Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 14/28] dyndbg: add ddebug_site(_get|_put) abstraction Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 15/28] dyndbg: ddebug_add_module avoid adding empty modules Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 16/28] dyndbg: add _index to struct _ddebug Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 17/28] dyndbg: prevent build bugs via -DNO_DYNAMIC_DEBUG_TABLE Jim Cromie
2021-05-12 14:55   ` Ard Biesheuvel
2021-05-12 18:03     ` jim.cromie
2021-05-11 18:50 ` [RFC PATCH v5 18/28] dyndbg: RFC - DEFINE_DYNAMIC_DEBUG_TABLE Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 19/28] dyndbg: RFC handle __dyndbg* sections in module.lds.h Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 20/28] dyndbg: ddebug_add_module() handle headers Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 21/28] dyndbg: validate ddebug_site_get invariants Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 22/28] dyndbg: fix NULL deref after deleting sites Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 23/28] dyndbg: dont show header records in control Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 24/28] dyndbg: make site pointer and checks on it optional (not quite) Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 25/28] dyndbg: swap WARN_ON for BUG_ON see what 0-day says Jim Cromie
2021-05-11 18:50 ` [RFC PATCH v5 26/28] dyndbg: fixup protect header when deleting site Jim Cromie
2021-05-11 18:50 ` Jim Cromie [this message]
2021-05-11 18:50 ` [RFC PATCH v5 28/28] dyndbg: RFC drop _ddebug.site pointer Jim Cromie

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210511185057.3815777-28-jim.cromie@gmail.com \
    --to=jim.cromie@gmail.com \
    --cc=jbaron@akamai.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox