* [PATCH] Make dcache multi-target-safe
@ 2021-09-24 15:50 Pedro Alves
2021-09-24 16:54 ` Andrew Burgess
0 siblings, 1 reply; 3+ messages in thread
From: Pedro Alves @ 2021-09-24 15:50 UTC (permalink / raw)
To: gdb-patches
By inspection, I noticed that this code in dcache.c is not
multi-target-aware:
/* If this is a different inferior from what we've recorded,
flush the cache. */
if (inferior_ptid != dcache->ptid)
This doesn't take into account that threads of different targets may
have the same ptid.
Fixed by also storing/comparing the process_stratum_target.
Tested on x86-64-linux-gnu, native and gdbserver.
Change-Id: I4d9d74052c696b72d28cb1c77b697b911725c8d3
---
gdb/dcache.c | 14 +++++++++++---
1 file changed, 11 insertions(+), 3 deletions(-)
diff --git a/gdb/dcache.c b/gdb/dcache.c
index 50b3854fe69..98365fee3b1 100644
--- a/gdb/dcache.c
+++ b/gdb/dcache.c
@@ -116,6 +116,10 @@ struct dcache_struct
/* The ptid of last inferior to use cache or null_ptid. */
ptid_t ptid;
+
+ /* The process target of last inferior to use the cache or
+ nullptr. */
+ process_stratum_target *proc_target;
};
typedef void (block_func) (struct dcache_block *block, void *param);
@@ -249,6 +253,7 @@ dcache_invalidate (DCACHE *dcache)
dcache->oldest = NULL;
dcache->size = 0;
dcache->ptid = null_ptid;
+ dcache->proc_target = nullptr;
if (dcache->line_size != dcache_line_size)
{
@@ -453,6 +458,7 @@ dcache_init (void)
dcache->size = 0;
dcache->line_size = dcache_line_size;
dcache->ptid = null_ptid;
+ dcache->proc_target = nullptr;
return dcache;
}
@@ -470,13 +476,15 @@ dcache_read_memory_partial (struct target_ops *ops, DCACHE *dcache,
{
ULONGEST i;
- /* If this is a different inferior from what we've recorded,
- flush the cache. */
+ /* If this is a different thread from what we've recorded, flush the
+ cache. */
- if (inferior_ptid != dcache->ptid)
+ process_stratum_target *proc_target = current_inferior ()->process_target ();
+ if (proc_target != dcache->proc_target || inferior_ptid != dcache->ptid)
{
dcache_invalidate (dcache);
dcache->ptid = inferior_ptid;
+ dcache->proc_target = proc_target;
}
for (i = 0; i < len; i++)
base-commit: 62df62b23032bb33ae7d7dc7ec7039a25794907b
--
2.26.2
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] Make dcache multi-target-safe
2021-09-24 15:50 [PATCH] Make dcache multi-target-safe Pedro Alves
@ 2021-09-24 16:54 ` Andrew Burgess
2021-09-24 17:55 ` Pedro Alves
0 siblings, 1 reply; 3+ messages in thread
From: Andrew Burgess @ 2021-09-24 16:54 UTC (permalink / raw)
To: Pedro Alves; +Cc: gdb-patches
* Pedro Alves <pedro@palves.net> [2021-09-24 16:50:11 +0100]:
> By inspection, I noticed that this code in dcache.c is not
> multi-target-aware:
>
> /* If this is a different inferior from what we've recorded,
> flush the cache. */
>
> if (inferior_ptid != dcache->ptid)
>
> This doesn't take into account that threads of different targets may
> have the same ptid.
>
> Fixed by also storing/comparing the process_stratum_target.
>
> Tested on x86-64-linux-gnu, native and gdbserver.
LGTM.
Thanks,
Andrew
>
> Change-Id: I4d9d74052c696b72d28cb1c77b697b911725c8d3
> ---
> gdb/dcache.c | 14 +++++++++++---
> 1 file changed, 11 insertions(+), 3 deletions(-)
>
> diff --git a/gdb/dcache.c b/gdb/dcache.c
> index 50b3854fe69..98365fee3b1 100644
> --- a/gdb/dcache.c
> +++ b/gdb/dcache.c
> @@ -116,6 +116,10 @@ struct dcache_struct
>
> /* The ptid of last inferior to use cache or null_ptid. */
> ptid_t ptid;
> +
> + /* The process target of last inferior to use the cache or
> + nullptr. */
> + process_stratum_target *proc_target;
> };
>
> typedef void (block_func) (struct dcache_block *block, void *param);
> @@ -249,6 +253,7 @@ dcache_invalidate (DCACHE *dcache)
> dcache->oldest = NULL;
> dcache->size = 0;
> dcache->ptid = null_ptid;
> + dcache->proc_target = nullptr;
>
> if (dcache->line_size != dcache_line_size)
> {
> @@ -453,6 +458,7 @@ dcache_init (void)
> dcache->size = 0;
> dcache->line_size = dcache_line_size;
> dcache->ptid = null_ptid;
> + dcache->proc_target = nullptr;
>
> return dcache;
> }
> @@ -470,13 +476,15 @@ dcache_read_memory_partial (struct target_ops *ops, DCACHE *dcache,
> {
> ULONGEST i;
>
> - /* If this is a different inferior from what we've recorded,
> - flush the cache. */
> + /* If this is a different thread from what we've recorded, flush the
> + cache. */
>
> - if (inferior_ptid != dcache->ptid)
> + process_stratum_target *proc_target = current_inferior ()->process_target ();
> + if (proc_target != dcache->proc_target || inferior_ptid != dcache->ptid)
> {
> dcache_invalidate (dcache);
> dcache->ptid = inferior_ptid;
> + dcache->proc_target = proc_target;
> }
>
> for (i = 0; i < len; i++)
>
> base-commit: 62df62b23032bb33ae7d7dc7ec7039a25794907b
> --
> 2.26.2
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] Make dcache multi-target-safe
2021-09-24 16:54 ` Andrew Burgess
@ 2021-09-24 17:55 ` Pedro Alves
0 siblings, 0 replies; 3+ messages in thread
From: Pedro Alves @ 2021-09-24 17:55 UTC (permalink / raw)
To: Andrew Burgess; +Cc: gdb-patches
On 2021-09-24 5:54 p.m., Andrew Burgess wrote:
> * Pedro Alves <pedro@palves.net> [2021-09-24 16:50:11 +0100]:
>
>> By inspection, I noticed that this code in dcache.c is not
>> multi-target-aware:
>>
>> /* If this is a different inferior from what we've recorded,
>> flush the cache. */
>>
>> if (inferior_ptid != dcache->ptid)
>>
>> This doesn't take into account that threads of different targets may
>> have the same ptid.
>>
>> Fixed by also storing/comparing the process_stratum_target.
>>
>> Tested on x86-64-linux-gnu, native and gdbserver.
>
> LGTM.
Thanks Andrew. Pushed.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2021-09-24 17:55 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-09-24 15:50 [PATCH] Make dcache multi-target-safe Pedro Alves
2021-09-24 16:54 ` Andrew Burgess
2021-09-24 17:55 ` Pedro Alves
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).