Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Tom de Vries <tdevries@suse.de>
To: "Metzger, Markus T" <markus.t.metzger@intel.com>
Cc: "gdb-patches@sourceware.org" <gdb-patches@sourceware.org>
Subject: Re: [PATCH][gdb/testsuite] Fix gdb.btrace/tsx.exp on system with tsx disabled in microcode
Date: Mon, 12 Jul 2021 16:05:48 +0200	[thread overview]
Message-ID: <0e162414-67d4-a05c-6603-c4521b480e1c@suse.de> (raw)
In-Reply-To: <DM8PR11MB5749C811C593FA625DA6DC13DE159@DM8PR11MB5749.namprd11.prod.outlook.com>

[-- Attachment #1: Type: text/plain, Size: 1982 bytes --]

On 7/12/21 1:54 PM, Metzger, Markus T wrote:
> Hello Tom,
> 
> 
>> diff --git a/gdb/testsuite/gdb.btrace/tsx.exp b/gdb/testsuite/gdb.btrace/tsx.exp
>> index ccde1ea807e..7f96313f1b1 100644
>> --- a/gdb/testsuite/gdb.btrace/tsx.exp
>> +++ b/gdb/testsuite/gdb.btrace/tsx.exp
>> @@ -59,6 +59,11 @@ set abort_2 [multi_line \
>>     "\[0-9\]*\t   0x\[0-9a-f\]+ <test\\+\[0-9\]+>:\tmov\[^\\\r\\\n\]*" \
>>     "\[0-9\]*\t   0x\[0-9a-f\]+ <test\\+\[0-9\]+>:\tret\[^\\\r\\\n\]*" \
>>     ]
>> +set abort_3 \
>> +    [multi_line \
>> +	 "$decimal\t   $hex <main\\+$decimal>:\tcall\[^\\\r\\\n\]*" \
>> +	 "$decimal\t   $hex <test\\+$decimal>:\tmov\[^\\\r\\\n\]*" \
>> +	 "$decimal\t   $hex <test\\+$decimal>:\tret\[^\\\r\\\n\]*"]
> 
> The patterns do not include the call since this is compiler-generated.  The actual
> TSX test is written in assembly so we know the instructions.
> 

I see.  Fixed by not requiring a specific instruction.

> 
>> set test "speculation indication"
>> gdb_test_multiple "record instruction-history" $test {
>> @@ -68,6 +73,9 @@ gdb_test_multiple "record instruction-history" $test {
>>     -re "$abort_2.*$gdb_prompt $" {
>>         pass $test
>>     }
>> +    -re -wrap "$abort_3" {
>> +        pass $gdb_test_name
>> +    }
> 
> Does this '-wrap' add ".*$gdb_prompt $"?
> 
> Note that we need the ".*" after the pattern since this code is compiler-generated
> and we don't really know when we will stop after returning from test ().


The -wrap processes the pattern in exactly the same way as gdb_test does:
...
        if { $wrap_pattern } {
            # Wrap subst_item as is done for the gdb_test PATTERN
            # argument.
            lappend $current_list \
                "\[\r\n\]*(?:$subst_item)\[\r\n\]+$gdb_prompt $"
            set wrap_pattern 0
        } else {
...

After running with clang, I ran into the FAIL you anticipated, which is
indeed fixed by adding .* in the pattern.

Updated patch attached.  OK for trunk?

Thanks,
- Tom


[-- Attachment #2: 0001-gdb-testsuite-Fix-gdb.btrace-tsx.exp-on-system-with-tsx-disabled-in-microcode.patch --]
[-- Type: text/x-patch, Size: 1925 bytes --]

[gdb/testsuite] Fix gdb.btrace/tsx.exp on system with tsx disabled in microcode

Recently I started to see this fail with trunk:
...
(gdb) record instruction-history^M
1          0x00000000004004ab <main+4>: call   0x4004b7 <test>^M
2          0x00000000004004c6 <test+15>:        mov    $0x1,%eax^M
3          0x00000000004004cb <test+20>:        ret    ^M
(gdb) FAIL: gdb.btrace/tsx.exp: speculation indication
...

This is due to an intel microcode update (1) that disables Intel TSX by default.

Fix this by updating the pattern.

Tested on x86_64-linux, with both gcc 7.5.0 and clang 12.0.1.

[1] https://www.intel.com/content/www/us/en/support/articles/000059422/processors.html

gdb/testsuite/ChangeLog:

2021-07-12  Tom de Vries  <tdevries@suse.de>

	PR testsuite/28057
	* gdb.btrace/tsx.exp: Add pattern for system with tsx disabled in
	microcode.

---
 gdb/testsuite/gdb.btrace/tsx.exp | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/gdb/testsuite/gdb.btrace/tsx.exp b/gdb/testsuite/gdb.btrace/tsx.exp
index ccde1ea807e..66f6305e50a 100644
--- a/gdb/testsuite/gdb.btrace/tsx.exp
+++ b/gdb/testsuite/gdb.btrace/tsx.exp
@@ -59,6 +59,11 @@ set abort_2 [multi_line \
     "\[0-9\]*\t   0x\[0-9a-f\]+ <test\\+\[0-9\]+>:\tmov\[^\\\r\\\n\]*" \
     "\[0-9\]*\t   0x\[0-9a-f\]+ <test\\+\[0-9\]+>:\tret\[^\\\r\\\n\]*" \
     ]
+set abort_3 \
+    [multi_line \
+	 "$decimal\t   $hex <main\\+$decimal>:\t\[^\\\r\\\n\]*" \
+	 "$decimal\t   $hex <test\\+$decimal>:\tmov\[^\\\r\\\n\]*" \
+	 "$decimal\t   $hex <test\\+$decimal>:\tret\[^\\\r\\\n\]*"]
 
 set test "speculation indication"
 gdb_test_multiple "record instruction-history" $test {
@@ -68,6 +73,9 @@ gdb_test_multiple "record instruction-history" $test {
     -re "$abort_2.*$gdb_prompt $" {
         pass $test
     }
+    -re -wrap "$abort_3.*" {
+        pass $gdb_test_name
+    }
     -re  "$begin_to_end.*$gdb_prompt $" {
         pass $test
     }

  reply	other threads:[~2021-07-12 14:08 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-12 11:27 Tom de Vries
2021-07-12 11:54 ` Metzger, Markus T via Gdb-patches
2021-07-12 14:05   ` Tom de Vries [this message]
2021-07-12 14:35     ` Metzger, Markus T via Gdb-patches
2021-07-12 15:08       ` Tom de Vries
2021-07-12 15:19         ` Metzger, Markus T via Gdb-patches

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=0e162414-67d4-a05c-6603-c4521b480e1c@suse.de \
    --to=tdevries@suse.de \
    --cc=gdb-patches@sourceware.org \
    --cc=markus.t.metzger@intel.com \
    /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