public inbox for libabigail@sourceware.org
 help / color / mirror / Atom feed
From: Matthias Maennich <maennich@google.com>
To: Giuliano Procida <gprocida@google.com>
Cc: libabigail@sourceware.org, dodji@seketeli.org, kernel-team@android.com
Subject: Re: [PATCH v2] abg-reader.cc: track WIP types by pointer not name
Date: Mon, 22 Jun 2020 22:06:30 +0200	[thread overview]
Message-ID: <20200622200630.GA75581@google.com> (raw)
In-Reply-To: <20200622172504.27660-1-gprocida@google.com>

On Mon, Jun 22, 2020 at 06:25:04PM +0100, Giuliano Procida wrote:
>When reading ABI XML files, the reader needs to construct types
>progressively as any type may depend on other types and even on
>itself. Such work-in-progress types are tracked explicitly.
>
>The storage used for this is a map from (external, qualified) type
>name to a count of how many times the type (name) has been seen.
>
>However, function type names are invariably stored as "void ()" as
>they are incomplete at the point they are added to the map. When the
>reader later attempts to remove the marking they have their proper,
>different names. In short, the code doesn't do what it's supposed to.
>
>This commit changes the stored value from string to const type_base*.
>Equality on type_base_sptr has been defined to be a deep comparison so
>storing those wouldn't be quite right.
>
>It also replaces the unordered_map with a simple vector, implementing
>a searchable stack. This is simpler and fast, for the expected number
>of types, and allows a stronger invariant to be asserted due to the
>preservation of insertion order.
>
>Note that this commit removes some of the call paths that result in
>incorrect (external) type names being cached, regardless of whether
>they are actually used.
>
>	* src/abg-reader.cc (xml_reader::m_wip_types_map): Replace
>	with m_wip_types_stack of type vector<const type_base*>.
>	(xml_reader::clear_wip_classes_map): Remove.
>	(xml_reader::clear_wip_types_stack): New function, clears
>	m_wip_types_stack.
>	(xml_reader::mark_type_as_wip): Push type pointer onto
>	m_wip_types_stack.
>	(xml_reader::unmark_type_as_wip): Add assertion that type
>	pointer is at top of stack. Pop pointer off m_wip_types_stack.
>	(xml_reader::is_wip_type): Test if type pointer in stack.
>

The surroundings of this code look like a bit of RAII could not harm.
Especially given the invariant that unmarking pops the stack, marking
could probably also define the unmarking by using something like a scope
guard.

Reviewed-by: Matthias Maennich <maennich@google.com>

Cheers,
Matthias

>Signed-off-by: Giuliano Procida <gprocida@google.com>
>---
> src/abg-reader.cc | 33 ++++++++++-----------------------
> 1 file changed, 10 insertions(+), 23 deletions(-)
>
>diff --git a/src/abg-reader.cc b/src/abg-reader.cc
>index eb74659f..480d1402 100644
>--- a/src/abg-reader.cc
>+++ b/src/abg-reader.cc
>@@ -26,6 +26,7 @@
> /// native XML format is named "abixml".
>
> #include "config.h"
>+#include <algorithm>
> #include <cstring>
> #include <cstdlib>
> #include <cerrno>
>@@ -119,7 +120,7 @@ private:
>   unordered_map<string, vector<type_base_sptr> >	m_types_map;
>   unordered_map<string, shared_ptr<function_tdecl> >	m_fn_tmpl_map;
>   unordered_map<string, shared_ptr<class_tdecl> >	m_class_tmpl_map;
>-  unordered_map<string, size_t>			m_wip_types_map;
>+  vector<const type_base*>				m_wip_types_stack;
>   vector<type_base_sptr>				m_types_to_canonicalize;
>   string_xml_node_map					m_id_xml_node_map;
>   xml_node_decl_base_sptr_map				m_xml_node_decl_map;
>@@ -522,8 +523,8 @@ public:
>   /// the map of the class that are currently being built, but at not
>   /// yet fully built.
>   void
>-  clear_wip_classes_map()
>-  {m_wip_types_map.clear();}
>+  clear_wip_types_stack()
>+  {m_wip_types_stack.clear();}
>
>   /// Mark a given type as being "Work In Progress"; that is, mark it
>   /// as being currently built.
>@@ -534,12 +535,7 @@ public:
>   {
>     if (!t)
>       return;
>-    string qname = get_type_name(t, /*qualified=*/true);
>-    unordered_map<string, size_t>::iterator it = m_wip_types_map.find(qname);
>-    if (it == m_wip_types_map.end())
>-      m_wip_types_map[qname] = 1;
>-    else
>-      ++it->second;
>+    m_wip_types_stack.push_back(t.get());
>   }
>
>   /// Mark a given class as being *NOT* "Work In Progress" anymore;
>@@ -551,15 +547,9 @@ public:
>   {
>     if (!t)
>       return;
>-
>-    string qname = get_type_name(t, /*qualified=*/true);
>-    unordered_map<string, size_t>::iterator it = m_wip_types_map.find(qname);
>-    if (it == m_wip_types_map.end())
>-      return;
>-    if (it->second)
>-      --it->second;
>-    if (it->second == 0)
>-      m_wip_types_map.erase(it);
>+    ABG_ASSERT(!m_wip_types_stack.empty());
>+    ABG_ASSERT(m_wip_types_stack.back() == t.get());
>+    m_wip_types_stack.pop_back();
>   }
>
>   /// Test if a type is being currently built; that is, if it's "Work
>@@ -571,11 +561,8 @@ public:
>   {
>     if (!t)
>       return false;
>-
>-    string qname = get_type_name(t, /*qualified=*/true);
>-    unordered_map<string, size_t>::const_iterator i =
>-      m_wip_types_map.find(qname);
>-    return i != m_wip_types_map.end();
>+    return std::find(m_wip_types_stack.begin(), m_wip_types_stack.end(), t.get())
>+           != m_wip_types_stack.end();
>   }
>
>   /// Test if two types are equal, without comparing them structurally.
>-- 
>2.27.0.111.gc72c7da667-goog
>

  reply	other threads:[~2020-06-22 20:06 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-22 16:06 [PATCH] " Giuliano Procida
2020-06-22 17:25 ` [PATCH v2] " Giuliano Procida
2020-06-22 20:06   ` Matthias Maennich [this message]
2020-07-09 12:58   ` Dodji Seketeli

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=20200622200630.GA75581@google.com \
    --to=maennich@google.com \
    --cc=dodji@seketeli.org \
    --cc=gprocida@google.com \
    --cc=kernel-team@android.com \
    --cc=libabigail@sourceware.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).