All of lore.kernel.org
 help / color / mirror / Atom feed
From: Catalin Marinas <catalin.marinas@arm.com>
To: linux-kernel@vger.kernel.org, linux-kbuild@vger.kernel.org
Cc: Sam Ravnborg <sam@ravnborg.org>
Subject: [RFC PATCH] kbuild: Do not select symbols with unmet dependencies
Date: Wed, 12 Aug 2009 12:39:40 +0100	[thread overview]
Message-ID: <20090812112308.30683.24700.stgit@pc1117.cambridge.arm.com> (raw)

The "select" statement in Kconfig files allows the enabling of options
even if they have unmet direct dependencies (i.e. "depends on" expands
to "no"). Currently, the "depends on" clauses are used in calculating
the visibility but they do not affect the reverse dependencies in any
way.

The patch introduces additional tracking of the "depends on" statements
and does not allow selecting an option if its direct dependencies are
not met, also printing a warning.

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
---

This patch is meant to make "select" safer in Kconfig files. The
"depends on", if not true, takes priority over "select".

For example, in -next, the ARM FRAME_POINTER (arch/arm/Kconfig.debug)
depends on !THUMB2_KERNEL because compiling the kernel to the Thumb-2
instruction set and FRAME_POINTER enabled fails. However, there are
several places where this symbol is forced by "select" statements (where
stacktrace support would be enough). Another example is DEBUG_PAGEALLOC
which depends on ARCH_SUPPORTS_DEBUG_PAGEALLOC (=n on ARM) but still
selected.

A safer approach currently may be to only warn but still select the
corresponding symbol.

The patch could probably be improved to also point to the symbols
containing the "select" statements, though I was getting some segfaults
with expr_list_for_each_sym(sym->rev_dep.expr).

Please note that the patch doesn't entirely follow the maximum line
length recommendation as none of the modified files comply with it.

Thank you for you comments.

 scripts/kconfig/expr.h   |    1 +
 scripts/kconfig/menu.c   |    5 +++++
 scripts/kconfig/symbol.c |   17 ++++++++++++++++-
 3 files changed, 22 insertions(+), 1 deletions(-)

diff --git a/scripts/kconfig/expr.h b/scripts/kconfig/expr.h
index 6408fef..2381562 100644
--- a/scripts/kconfig/expr.h
+++ b/scripts/kconfig/expr.h
@@ -83,6 +83,7 @@ struct symbol {
 	tristate visible;
 	int flags;
 	struct property *prop;
+	struct expr_value dep;
 	struct expr_value rev_dep;
 };
 
diff --git a/scripts/kconfig/menu.c b/scripts/kconfig/menu.c
index 07ff8d1..7d32d74 100644
--- a/scripts/kconfig/menu.c
+++ b/scripts/kconfig/menu.c
@@ -385,6 +385,11 @@ void menu_finalize(struct menu *parent)
 				expr_alloc_and(parent->prompt->visible.expr,
 					expr_alloc_symbol(&symbol_mod)));
 	}
+
+	if (sym) {
+		sym->dep.expr = expr_transform(expr_copy(parent->dep));
+		sym->dep.expr = expr_eliminate_dups(sym->dep.expr);
+	}
 }
 
 bool menu_is_visible(struct menu *menu)
diff --git a/scripts/kconfig/symbol.c b/scripts/kconfig/symbol.c
index 18f3e5c..af362d0 100644
--- a/scripts/kconfig/symbol.c
+++ b/scripts/kconfig/symbol.c
@@ -205,6 +205,16 @@ static void sym_calc_visibility(struct symbol *sym)
 	}
 	if (sym_is_choice_value(sym))
 		return;
+	/* defaulting to "yes" if no explicit "depends on" are given */
+	tri = yes;
+	if (sym->dep.expr)
+		tri = expr_calc_value(sym->dep.expr);
+	if (tri == mod)
+		tri = yes;
+	if (sym->dep.tri != tri) {
+		sym->dep.tri = tri;
+		sym_set_changed(sym);
+	}
 	tri = no;
 	if (sym->rev_dep.expr)
 		tri = expr_calc_value(sym->rev_dep.expr);
@@ -321,7 +331,12 @@ void sym_calc_value(struct symbol *sym)
 				}
 			}
 		calc_newval:
-			newval.tri = EXPR_OR(newval.tri, sym->rev_dep.tri);
+			if (sym->dep.tri == no && sym->rev_dep.tri != no)
+				fprintf(stderr, "%s:%d:warning: not selecting symbol with unmet dependencies: %s\n",
+					sym->prop->file->name, sym->prop->lineno,
+					sym->name ? sym->name : "<choice>");
+			else
+				newval.tri = EXPR_OR(newval.tri, sym->rev_dep.tri);
 		}
 		if (newval.tri == mod && sym_get_type(sym) == S_BOOLEAN)
 			newval.tri = yes;


             reply	other threads:[~2009-08-12 11:39 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-08-12 11:39 Catalin Marinas [this message]
2009-08-12 12:36 ` [RFC PATCH] kbuild: Do not select symbols with unmet dependencies Arnd Bergmann
2009-08-12 14:00   ` Catalin Marinas
2009-08-12 15:55     ` Catalin Marinas

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=20090812112308.30683.24700.stgit@pc1117.cambridge.arm.com \
    --to=catalin.marinas@arm.com \
    --cc=linux-kbuild@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=sam@ravnborg.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.