Looping back the mailing-lists.
On Thu, Oct 3, 2013 at 4:23 PM, Fabien Parent <fparent@xxxxxxxxxxxx> wrote:
>
> Hi David,
>
>
> On Wed, Oct 2, 2013 at 2:59 PM, David Gibson <david@xxxxxxxxxxxxxxxxxxxxx> wrote:
>>
>> On Tue, Sep 24, 2013 at 06:52:07PM +0200, Benoit Cousson wrote:
>> > From: Fabien Parent <fparent@xxxxxxxxxxxx>
>>
>> As noted elsewhere, please write patches against upstream dtc, not the
>> version in the kernel. git://git.kernel.org/pub/scm/utils/dtc/dtc.git
>>
>> > There are a few memory leaks in dtc which until now were not that important
>> > since they were all in the parser and only one instance of the parser was run
>> > per instance of dtc. The following commits will add a validation of dts through
>> > schema which have the same syntax as dts, i.e. the parser of dts will be reused
>> > to parse schema. The consequence is that instead of having the parser running
>> > only one time for an instance of the dtc process, the parser will run
>> > many many times and thus the leak that were not important until now becomes
>> > urgent to fix.
>>
>> Hrm, yeah. I'm aware that I haven't been very careful with memory
>> leaks within the parser. Essentially, I've been treating the process
>> as a pool allocator with exactly one pool - I've even considered
>> getting rid of those free()s we do have.
>>
>> If the parser's going to be invoked repeatedly, then, yes, that
>> certainly needs fixing. Whether individually fixing each leak or
>> using a explicit pool allocator is a better option is less clear.
>
>
> I've chosen the path of using free()s since it was removing most (all?) memory leaks from
> the parser and wasn't adding much more complexity to it. I don't think a pool allocator would
> be useful to DTC in its current state, but it's true that with schemas which are using
> the DTS syntax it could make a lot of sense to switch to a pool allocator.
>
> I guess I will wait until the discussion about this proposal has progressed and see
> whether this patch should be rewritten using a pool allocator or not.
>
>>
>> > dtc-lexer: Do not duplicate the string which contains literals because the
>> > string is directly converted afterward to an integer and is never used again.
>> > livetree: Add a bunch of free helper functions to clean properly the
>> > dt.
>>
>> This is no good. Making this assumption is a layering violation - if
>> the parser was changed so that this literal wasn't converted until
>> after another token was read, the yytext value copied in here would no
>> longer be valid and would break horribly.
>>
>
> Right, I've been lazy on this one and I took the easy road.
>
>
>>
>> >
>> > Signed-off-by: Fabien Parent <fparent@xxxxxxxxxxxx>
>> > Signed-off-by: Benoit Cousson <bcousson@xxxxxxxxxxxx>
>> > ---
>> > scripts/dtc/dtc-lexer.l | 2 +-
>> > scripts/dtc/dtc-lexer.lex.c_shipped | 2 +-
>> > scripts/dtc/dtc.c | 1 +
>> > scripts/dtc/dtc.h | 1 +
>> > scripts/dtc/livetree.c | 108 +++++++++++++++++++++++++++++++++---
>> > 5 files changed, 105 insertions(+), 9 deletions(-)
>> >
>> > diff --git a/scripts/dtc/dtc-lexer.l b/scripts/dtc/dtc-lexer.l
>> > index 3b41bfc..4f63fbf 100644
>> > --- a/scripts/dtc/dtc-lexer.l
>> > +++ b/scripts/dtc/dtc-lexer.l
>> > @@ -146,7 +146,7 @@ static int pop_input_file(void);
>> > }
>> >
>> > <V1>([0-9]+|0[xX][0-9a-fA-F]+)(U|L|UL|LL|ULL)? {
>> > - yylval.literal = xstrdup(yytext);
>> > + yylval.literal = yytext;
>> > DPRINT("Literal: '%s'\n", yylval.literal);
>> > return DT_LITERAL;
>> > }
>> > diff --git a/scripts/dtc/dtc-lexer.lex.c_shipped b/scripts/dtc/dtc-lexer.lex.c_shipped
>> > index 2d30f41..5c0d27c 100644
>> > --- a/scripts/dtc/dtc-lexer.lex.c_shipped
>> > +++ b/scripts/dtc/dtc-lexer.lex.c_shipped
>> > @@ -1054,7 +1054,7 @@ case 10:
>> > YY_RULE_SETUP
>> > #line 148 "dtc-lexer.l"
>> > {
>> > - yylval.literal = xstrdup(yytext);
>> > + yylval.literal = yytext;
>> > DPRINT("Literal: '%s'\n", yylval.literal);
>> > return DT_LITERAL;
>> > }
>> > diff --git a/scripts/dtc/dtc.c b/scripts/dtc/dtc.c
>> > index a375683..215ae92 100644
>> > --- a/scripts/dtc/dtc.c
>> > +++ b/scripts/dtc/dtc.c
>> > @@ -256,5 +256,6 @@ int main(int argc, char *argv[])
>> > die("Unknown output format \"%s\"\n", outform);
>> > }
>> >
>> > + free_dt(bi);
>> > exit(0);
>> > }
>> > diff --git a/scripts/dtc/dtc.h b/scripts/dtc/dtc.h
>> > index 3e42a07..9c45fd2 100644
>> > --- a/scripts/dtc/dtc.h
>> > +++ b/scripts/dtc/dtc.h
>> > @@ -245,6 +245,7 @@ struct boot_info {
>> > struct boot_info *build_boot_info(struct reserve_info *reservelist,
>> > struct node *tree, uint32_t boot_cpuid_phys);
>> > void sort_tree(struct boot_info *bi);
>> > +void free_dt(struct boot_info *bi);
>> >
>> > /* Checks */
>> >
>> > diff --git a/scripts/dtc/livetree.c b/scripts/dtc/livetree.c
>> > index b61465f..5c8692c 100644
>> > --- a/scripts/dtc/livetree.c
>> > +++ b/scripts/dtc/livetree.c
>> > @@ -20,6 +20,10 @@
>> >
>> > #include "dtc.h"
>> >
>> > +static void free_node_list(struct node *n);
>> > +static void free_node(struct node *n);
>> > +static void free_property(struct property *p);
>> > +
>> > /*
>> > * Tree building functions
>> > */
>> > @@ -144,7 +148,7 @@ struct node *merge_nodes(struct node *old_node, struct node *new_node)
>> >
>> > /* Add new node labels to old node */
>> > for_each_label_withdel(new_node->labels, l)
>> > - add_label(&old_node->labels, l->label);
>> > + add_label(&old_node->labels, xstrdup(l->label));
>> >
>> > /* Move properties from the new node to the old node. If there
>> > * is a collision, replace the old value with the new */
>> > @@ -156,7 +160,7 @@ struct node *merge_nodes(struct node *old_node, struct node *new_node)
>> >
>> > if (new_prop->deleted) {
>> > delete_property_by_name(old_node, new_prop->name);
>> > - free(new_prop);
>> > + free_property(new_prop);
>> > continue;
>> > }
>> >
>> > @@ -165,7 +169,7 @@ struct node *merge_nodes(struct node *old_node, struct node *new_node)
>> > if (streq(old_prop->name, new_prop->name)) {
>> > /* Add new labels to old property */
>> > for_each_label_withdel(new_prop->labels, l)
>> > - add_label(&old_prop->labels, l->label);
>> > + add_label(&old_prop->labels, xstrdup(l->label));
>> >
>> > old_prop->val = new_prop->val;
>> > old_prop->deleted = 0;
>> > @@ -191,7 +195,7 @@ struct node *merge_nodes(struct node *old_node, struct node *new_node)
>> >
>> > if (new_child->deleted) {
>> > delete_node_by_name(old_node, new_child->name);
>> > - free(new_child);
>> > + free_node(new_child);
>> > continue;
>> > }
>> >
>> > @@ -211,7 +215,7 @@ struct node *merge_nodes(struct node *old_node, struct node *new_node)
>> >
>> > /* The new node contents are now merged into the old node. Free
>> > * the new node. */
>> > - free(new_node);
>> > + free_node(new_node);
>> >
>> > return old_node;
>> > }
>> > @@ -532,13 +536,13 @@ cell_t get_node_phandle(struct node *root, struct node *node)
>> > if (!get_property(node, "linux,phandle")
>> > && (phandle_format & PHANDLE_LEGACY))
>> > add_property(node,
>> > - build_property("linux,phandle",
>> > + build_property(xstrdup("linux,phandle"),
>> > data_append_cell(empty_data, phandle)));
>> >
>> > if (!get_property(node, "phandle")
>> > && (phandle_format & PHANDLE_EPAPR))
>> > add_property(node,
>> > - build_property("phandle",
>> > + build_property(xstrdup("phandle"),
>> > data_append_cell(empty_data, phandle)));
>> >
>> > /* If the node *does* have a phandle property, we must
>> > @@ -707,3 +711,93 @@ void sort_tree(struct boot_info *bi)
>> > sort_reserve_entries(bi);
>> > sort_node(bi->dt);
>> > }
>> > +
>> > +static void free_marker_list(struct marker *m)
>> > +{
>> > + struct marker *marker, *marker_next;
>> > +
>> > + if (!m)
>> > + return;
>> > +
>> > + for (marker = m, marker_next = marker ? marker->next : NULL;
>> > + marker;
>> > + marker = marker_next, marker_next = marker ? marker->next : NULL) {
>>
>> Rather hard to read version of safe-against-free list iteration.
>>
>
> Agreed.
>
>>
>> > + free(marker->ref);
>> > + free(marker);
>> > + }
>> > +}
>> > +
>> > +static void free_label_list(struct label *l)
>> > +{
>> > + struct label *label, *label_next;
>> > +
>> > + if (!l)
>> > + return;
>> > +
>> > + for (label = l, label_next = label ? label->next : NULL;
>> > + label;
>> > + label = label_next, label_next = label ? label->next : NULL) {
>> > + free(label->label);
>> > + free(label);
>> > + }
>> > +}
>> > +
>> > +static void free_property(struct property *p)
>> > +{
>> > + if (!p)
>> > + return;
>> > +
>> > + free_label_list(p->labels);
>> > + free_marker_list(p->val.markers);
>> > + free(p->val.val);
>> > + free(p->name);
>> > + free(p);
>> > +}
>> > +
>> > +static void free_property_list(struct property *p)
>> > +{
>> > + struct property *next;
>> > +
>> > + if (!p)
>> > + return;
>> > +
>> > + for (next = p->next; p; p = next, next = p ? p->next : NULL)
>> > + free_property(p);
>> > +}
>> > +
>> > +static void free_node(struct node *n)
>> > +{
>> > + if (!n)
>> > + return;
>> > +
>> > + free_node_list(n->children);
>> > + free_label_list(n->labels);
>> > + free_property_list(n->proplist);
>> > + free(n->fullpath);
>> > + if (n->name && *n->name)
>>
>> *n->name .. so, the name can (and must) be statically allocated if
>> it's exactly "". Not a useful optimization, I think.
>
>
> True.
>
>>
>>
>> > + free(n->name);
>> > + free(n);
>> > +}
>> > +
>> > +static void free_node_list(struct node *n)
>> > +{
>> > + struct node *next;
>> > +
>> > + if (!n)
>> > + return;
>> > +
>> > + for (next = n->next_sibling;
>> > + n;
>> > + n = next, next = n ? n->next_sibling : NULL) {
>> > + free_node(n);
>> > + }
>> > +}
>> > +
>> > +void free_dt(struct boot_info *bi)
>> > +{
>> > + if (!bi)
>> > + return;
>> > +
>> > + free_node_list(bi->dt);
>> > + free(bi);
>> > +}
>>
>> --
>> David Gibson | I'll have my music baroque, and my code
>> david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
>> | _way_ _around_!
>> http://www.ozlabs.org/~dgibson
>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
