public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Simon Marchi <simon.marchi@efficios.com>
To: gdb-patches@sourceware.org
Cc: Simon Marchi <simon.marchi@efficios.com>
Subject: [PATCH 4/9] gdb: add type parameter to value::allocate_register and add value::allocate_register_lazy
Date: Thu, 21 Dec 2023 14:16:25 -0500	[thread overview]
Message-ID: <20231221191716.257256-5-simon.marchi@efficios.com> (raw)
In-Reply-To: <20231221191716.257256-1-simon.marchi@efficios.com>

Some places that create register struct values don't use register_type
to obtain the value type.  This prevents them from using the current
version of value::allocate_register.  One spot (value_of_register_lazy)
also creates a lazy register value.

Add a value::allocate_register_lazy method.  Add some type parameters
to value::allocate_register and value::allocate_register_lazy, to let
the caller specify the type to use for the value.  The parameters
default to nullptr, in which case we use register_type to obtain the
type.

Change-Id: I640ec0a5a0f4a55eba12d515dbfd25933229f8ec
---
 gdb/findvar.c     | 24 ++++++------------------
 gdb/rs6000-tdep.c | 13 ++++---------
 gdb/value.c       | 22 ++++++++++++++++++----
 gdb/value.h       | 13 +++++++++----
 4 files changed, 37 insertions(+), 35 deletions(-)

diff --git a/gdb/findvar.c b/gdb/findvar.c
index 852f42dc176d..7c6c837ae12e 100644
--- a/gdb/findvar.c
+++ b/gdb/findvar.c
@@ -281,12 +281,7 @@ value_of_register_lazy (frame_info_ptr next_frame, int regnum)
   /* We should have a valid next frame.  */
   gdb_assert (frame_id_p (get_frame_id (next_frame)));
 
-  value *reg_val = value::allocate_lazy (register_type (gdbarch, regnum));
-  reg_val->set_lval (lval_register);
-  VALUE_REGNUM (reg_val) = regnum;
-  VALUE_NEXT_FRAME_ID (reg_val) = get_frame_id (next_frame);
-
-  return reg_val;
+  return value::allocate_register_lazy (next_frame, regnum);
 }
 
 /* Given a pointer of type TYPE in target form in BUF, return the
@@ -751,25 +746,20 @@ value *
 default_value_from_register (gdbarch *gdbarch, type *type, int regnum,
 			     frame_info_ptr this_frame)
 {
-  int len = type->length ();
-  struct value *value = value::allocate (type);
-  value->set_lval (lval_register);
-
   frame_info_ptr next_frame = get_next_frame_sentinel_okay (this_frame);
   while (get_frame_type (next_frame) == INLINE_FRAME)
     next_frame = get_next_frame_sentinel_okay (next_frame);
 
-  VALUE_NEXT_FRAME_ID (value) = get_frame_id (next_frame);
-  VALUE_REGNUM (value) = regnum;
+  value *value = value::allocate_register (next_frame, regnum, type);
 
   /* Any structure stored in more than one register will always be
      an integral number of registers.  Otherwise, you need to do
      some fiddling with the last register copied here for little
      endian machines.  */
   if (type_byte_order (type) == BFD_ENDIAN_BIG
-      && len < register_size (gdbarch, regnum))
+      && type->length () < register_size (gdbarch, regnum))
     /* Big-endian, and we want less than full size.  */
-    value->set_offset (register_size (gdbarch, regnum) - len);
+    value->set_offset (register_size (gdbarch, regnum) - type->length ());
   else
     value->set_offset (0);
 
@@ -842,10 +832,8 @@ value_from_register (struct type *type, int regnum, frame_info_ptr frame)
 	 the corresponding [integer] type (see Alpha).  The assumption
 	 is that gdbarch_register_to_value populates the entire value
 	 including the location.  */
-      v = value::allocate (type);
-      v->set_lval (lval_register);
-      VALUE_NEXT_FRAME_ID (v) = get_frame_id (get_next_frame_sentinel_okay (frame));
-      VALUE_REGNUM (v) = regnum;
+      v = value::allocate_register (get_next_frame_sentinel_okay (frame),
+				    regnum, type);
       ok = gdbarch_register_to_value (gdbarch, frame, regnum, type1,
 				      v->contents_raw ().data (), &optim,
 				      &unavail);
diff --git a/gdb/rs6000-tdep.c b/gdb/rs6000-tdep.c
index 478e9078de6f..edf776853e20 100644
--- a/gdb/rs6000-tdep.c
+++ b/gdb/rs6000-tdep.c
@@ -2751,30 +2751,25 @@ static value *
 rs6000_value_from_register (gdbarch *gdbarch, type *type, int regnum,
 			    frame_info_ptr this_frame)
 {
-  int len = type->length ();
-  struct value *value = value::allocate (type);
-
   /* We have an IEEE 128-bit float -- need to change regnum mapping from
      fpr to vsr.  */
   regnum = ieee_128_float_regnum_adjust (gdbarch, type, regnum);
 
-  value->set_lval (lval_register);
-
   frame_info_ptr next_frame = get_next_frame_sentinel_okay (this_frame);
   while (get_frame_type (next_frame) == INLINE_FRAME)
     next_frame = get_next_frame_sentinel_okay (next_frame);
 
-  VALUE_NEXT_FRAME_ID (value) = get_frame_id (next_frame);
-  VALUE_REGNUM (value) = regnum;
+  value *value
+    = value::allocate_register (next_frame, regnum, type);
 
   /* Any structure stored in more than one register will always be
      an integral number of registers.  Otherwise, you need to do
      some fiddling with the last register copied here for little
      endian machines.  */
   if (type_byte_order (type) == BFD_ENDIAN_BIG
-      && len < register_size (gdbarch, regnum))
+      && type->length () < register_size (gdbarch, regnum))
     /* Big-endian, and we want less than full size.  */
-    value->set_offset (register_size (gdbarch, regnum) - len);
+    value->set_offset (register_size (gdbarch, regnum) - type->length ());
   else
     value->set_offset (0);
 
diff --git a/gdb/value.c b/gdb/value.c
index 7523af142348..7d51396a0e3a 100644
--- a/gdb/value.c
+++ b/gdb/value.c
@@ -961,11 +961,14 @@ value::allocate (struct type *type)
 
 /* See value.h  */
 
-struct value *
-value::allocate_register (frame_info_ptr next_frame, int regnum)
+value *
+value::allocate_register_lazy (frame_info_ptr next_frame, int regnum,
+			       struct type *type)
 {
-  value *result
-    = value::allocate (register_type (frame_unwind_arch (next_frame), regnum));
+  if (type == nullptr)
+    type = register_type (frame_unwind_arch (next_frame), regnum);
+
+  value *result = value::allocate_lazy (type);
 
   result->set_lval (lval_register);
   VALUE_REGNUM (result) = regnum;
@@ -974,6 +977,17 @@ value::allocate_register (frame_info_ptr next_frame, int regnum)
   return result;
 }
 
+/* See value.h  */
+
+value *
+value::allocate_register (frame_info_ptr next_frame, int regnum,
+			  struct type *type)
+{
+  value *result = value::allocate_register_lazy (next_frame, regnum, type);
+  result->set_lazy (false);
+  return result;
+}
+
 /* Allocate a  value  that has the correct length
    for COUNT repetitions of type TYPE.  */
 
diff --git a/gdb/value.h b/gdb/value.h
index d4244dadb91a..78abcac7e739 100644
--- a/gdb/value.h
+++ b/gdb/value.h
@@ -159,12 +159,17 @@ struct value
   /* Allocate a value and its contents for type TYPE.  */
   static struct value *allocate (struct type *type);
 
-  /* Allocate a non-lazy value representing register RENUM in the frame previous
-     to NEXT_FRAME.  The type of the value is found using `register_type`.
-
+  /* Allocate a lazy value representing register REGNUM in the frame previous
+     to NEXT_FRAME.  If TYPE is non-nullptr, use it as the value type.
+     Otherwise, use `register_type` to obtain the type.  */
+  static struct value *allocate_register_lazy (frame_info_ptr next_frame,
+					  int regnum, type *type = nullptr);
+
+  /* Same as `allocate_register_lazy`, but make the value non-lazy.
+  
      The caller is responsible for filling the value's contents.  */
   static struct value *allocate_register (frame_info_ptr next_frame,
-					  int regnum);
+					  int regnum, type *type = nullptr);
 
   /* Create a computed lvalue, with type TYPE, function pointers
      FUNCS, and closure CLOSURE.  */
-- 
2.43.0


  parent reply	other threads:[~2023-12-21 19:17 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-21 19:16 [PATCH 0/9] Some register value cleanups Simon Marchi
2023-12-21 19:16 ` [PATCH 1/9] gdb: don't set frame id after calling cooked_read_value Simon Marchi
2023-12-21 19:16 ` [PATCH 2/9] gdb: pass frame_info_ptr to gdbarch_value_from_register Simon Marchi
2023-12-22 16:40   ` Tom Tromey
2023-12-22 16:53     ` Simon Marchi
2023-12-22 16:56       ` Tom Tromey
2023-12-21 19:16 ` [PATCH 3/9] gdb: pass non-nullptr frame to gdbarch_value_from_register in address_from_register Simon Marchi
2023-12-21 19:16 ` Simon Marchi [this message]
2023-12-21 19:16 ` [PATCH 5/9] gdb: remove read_frame_register_value's frame parameter Simon Marchi
2023-12-21 19:16 ` [PATCH 6/9] gdb: implement address_from_register using value_from_register Simon Marchi
2023-12-21 19:16 ` [PATCH 7/9] gdb: remove VALUE_NEXT_FRAME_ID, add value::next_frame_id Simon Marchi
2023-12-22 16:51   ` Tom Tromey
2023-12-22 16:56     ` Simon Marchi
2023-12-22 17:02       ` Tom Tromey
2023-12-22 17:06         ` Simon Marchi
2023-12-24 15:35           ` Simon Marchi
2023-12-21 19:16 ` [PATCH 8/9] gdb: remove VALUE_REGNUM, add value::regnum Simon Marchi
2023-12-22 16:52   ` Tom Tromey
2023-12-22 16:57     ` Simon Marchi
2023-12-21 19:16 ` [PATCH 9/9] gdb: make value::allocate_register_lazy store id of next non-inline frame Simon Marchi
2023-12-22 16:53 ` [PATCH 0/9] Some register value cleanups Tom Tromey
2023-12-22 16:58   ` Simon Marchi
2023-12-22 17:02     ` Tom Tromey
2023-12-24 18:28       ` Simon Marchi

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=20231221191716.257256-5-simon.marchi@efficios.com \
    --to=simon.marchi@efficios.com \
    --cc=gdb-patches@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).