Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Yao Qi <yao@codesourcery.com>
To: <gdb-patches@sourceware.org>
Cc: Wei-cheng Wang <cole945@gmail.com>
Subject: Re: [PATCH] Fix memory-region overlapping checking
Date: Wed, 08 Aug 2012 10:31:00 -0000	[thread overview]
Message-ID: <1450628.xUgDUToN8c@qiyao.dyndns.org> (raw)
In-Reply-To: <CAPmZyH65a8B0=9wwMVHhDC-eTxOQ1==MRhhcFP=bDuZCJa9tBQ@mail.gmail.com>

Wei-cheng,
Thanks for writing a test.  Some comments below,

On Tuesday, August 07, 2012 02:01:59 PM Wei-cheng Wang wrote:
> +
> +#
> +# Test overlapping checking
> +#
> +#       lo'        hi'
> +#        |---------|
> +#  10 20 30 40 50 60 70 80
> +#      |----|             FAIL
> +#           |---|         FAIL
> +#               |----|    FAIL
> +#      |-------------|    FAIL
> +#   |----|                PASS
> +#                  |----| PASS
> +
> +proc delete_memory {} {
> +    global gdb_prompt
> +
> +    send_gdb "delete mem\n"
> +    gdb_expect 100 {
> +       -re "Delete all memory regions.*y or n.*$" {
> +           send_gdb "y\n";
> +           exp_continue
> +       }
> +       -re "$gdb_prompt $" { }
> +    }

We can use gdb_test_multiple here.

> +}
> +
> +# Test normal case (upper != 0)
> +delete_memory
> +send_gdb "mem 0x30 0x60 ro\n"; gdb_expect -re "$gdb_prompt $"

We can use gdb_test_no_output "mem 0x30 0x60 ro"

> +gdb_test "mem 0x20 0x40 ro" "overlapping memory region"
> +gdb_test "mem 0x40 0x50 ro" "overlapping memory region"
> +gdb_test "mem 0x50 0x70 ro" "overlapping memory region"
> +gdb_test_no_output "mem 0x10 0x30 ro"
> +gdb_test_no_output "mem 0x60 0x80 ro"
> +
> +# Test sepcial case (upper == 0)
          ^^^^^^^ typo "special"

> +delete_memory
> +send_gdb "mem 0x30 0x0 ro\n"; gdb_expect -re "$gdb_prompt $"

gdb_test_no_output

> +gdb_test "mem 0x20 0x40 ro" "overlapping memory region"
> +gdb_test "mem 0x40 0x50 ro" "overlapping memory region"
> +gdb_test "mem 0x50 0x70 ro" "overlapping memory region"
> +gdb_test_no_output "mem 0x10 0x30 ro"

the test result in summary is *not* unique,
$ cat testsuite/gdb.sum  | grep "PASS" | sort | uniq -c | sort -n
....
      2 PASS: gdb.base/memattr.exp: mem 0x20 0x40 ro
      2 PASS: gdb.base/memattr.exp: mem 0x40 0x50 ro
      2 PASS: gdb.base/memattr.exp: mem 0x50 0x70 ro

See more about this rule here 
http://sourceware.org/gdb/wiki/GDBTestcaseCookbook#Make_sure_test_messages_are_unique

Below is a new one fixed all the issues in your original patch.  What do you 
think?

Without your fix, the new test exposes a fail,

  FAIL: gdb.base/memattr.exp: 0x10 0x30 2

and fail is fixed with the patch applied.

-- 
Yao (齐尧)

2012-08-08  Wei-cheng Wang  <cole945@gmail.com>
		 Yao Qi <yao@codesourcery.com>

        * gdb.base/memattr.exp: Add cases for overlapping checking.

diff --git INDEX:/gdb/testsuite/gdb.base/memattr.exp 
--- INDEX:/gdb/testsuite/gdb.base/memattr.exp
+++ WORKDIR:/gdb/testsuite/gdb.base/memattr.exp
@@ -448,3 +448,45 @@ gdb_test_multiple "info mem" "mem 2-4 were deleted" {
 
 gdb_test "delete mem 8" "No memory region number 8." \
     "delete non-existant region"
+
+#
+# Test overlapping checking
+#
+#       lo'        hi'
+#        |---------|
+#  10 20 30 40 50 60 70 80
+#      |----|             FAIL
+#           |---|         FAIL
+#               |----|    FAIL
+#      |-------------|    FAIL
+#   |----|                PASS
+#                  |----| PASS
+
+proc delete_memory {} {
+    global gdb_prompt
+
+    gdb_test_multiple "delete mem" "delete mem" {
+	-re "Delete all memory regions.*y or n.*$" {
+	    send_gdb "y\n";
+	    exp_continue
+	}
+	-re "$gdb_prompt $" { }
+    }
+}
+
+# Test normal case (upper != 0)
+delete_memory
+gdb_test_no_output "mem 0x30 0x60 ro"
+gdb_test "mem 0x20 0x40 ro" "overlapping memory region" "0x20 0x40 1"
+gdb_test "mem 0x40 0x50 ro" "overlapping memory region" "0x40 0x50 1"
+gdb_test "mem 0x50 0x70 ro" "overlapping memory region" "0x50 0x70 1"
+gdb_test_no_output "mem 0x10 0x30 ro" "0x10 0x30 1"
+gdb_test_no_output "mem 0x60 0x80 ro"
+
+# Test special case (upper == 0)
+delete_memory
+gdb_test_no_output "mem 0x30 0x0 ro"
+gdb_test "mem 0x20 0x40 ro" "overlapping memory region" "0x20 0x40 2"
+gdb_test "mem 0x40 0x50 ro" "overlapping memory region" "0x40 0x50 2"
+gdb_test "mem 0x50 0x70 ro" "overlapping memory region" "0x50 0x70 2"
+gdb_test_no_output "mem 0x10 0x30 ro" "0x10 0x30 2"


  reply	other threads:[~2012-08-08 10:31 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-08-02  6:11 Wei-cheng Wang
2012-08-02  8:10 ` Yao Qi
2012-08-02 10:00   ` Wei-cheng Wang
2012-08-03 10:54 ` Yao Qi
2012-08-07  2:35   ` Wei-cheng Wang
2012-08-07  2:48     ` Yao Qi
2012-08-07  6:02       ` Wei-cheng Wang
2012-08-08 10:31         ` Yao Qi [this message]
2012-08-16 10:03           ` Wei-cheng Wang
2012-08-24 11:18             ` Pedro Alves
2012-08-24 11:47               ` Wei-cheng Wang
2012-08-24 15:12                 ` Pedro Alves
2012-08-27  2:22                   ` Wei-cheng Wang
2012-08-27  8:58                     ` Pedro Alves

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=1450628.xUgDUToN8c@qiyao.dyndns.org \
    --to=yao@codesourcery.com \
    --cc=cole945@gmail.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