linux-kernel - Re: [PATCH v2 2/2] mm/zbud: don't export any zbud API

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <813ac556-4a48-1e29-ad87-1ddb74546ef1@huawei.com>
Date:   Fri, 18 Jun 2021 10:28:56 +0800
From:   Miaohe Lin <linmiaohe@...wei.com>
To:     Nathan Chancellor <nathan@...nel.org>
CC:     <akpm@...ux-foundation.org>, <sjenning@...hat.com>,
        <ddstreet@...e.org>, <linux-kernel@...r.kernel.org>,
        <linux-mm@...ck.org>
Subject: Re: [PATCH v2 2/2] mm/zbud: don't export any zbud API

Hi:
On 2021/6/18 8:44, Nathan Chancellor wrote:
> On Tue, Jun 08, 2021 at 07:45:15PM +0800, Miaohe Lin wrote:
>> The zbud doesn't need to export any API and it is meant to be used via
>> zpool API since the commit 12d79d64bfd3 ("mm/zpool: update zswap to use
>> zpool"). So we can remove the unneeded zbud.h and move down zpool API
>> to avoid any forward declaration.
>>
>> Signed-off-by: Miaohe Lin <linmiaohe@...wei.com>
> 

Thank you for help figure these warnings out. It seems zbud module won't do
anything when CONFIG_ZPOOL is disabled. I think we should make zbud depends
on ZPOOL and eliminate the CONFIG_ZPOOL macro in zbud.c like what z3fold does.
Does this make sense for you?
Thanks again. :)

diff --git a/mm/Kconfig b/mm/Kconfig
index 8f748010f7ea..5dc28e9205e0 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -674,6 +674,7 @@ config ZPOOL

 config ZBUD
        tristate "Low (Up to 2x) density storage for compressed pages"
+       depends on ZPOOL
        help
          A special purpose allocator for storing compressed pages.
          It is designed to store up to two compressed pages per physical
diff --git a/mm/zbud.c b/mm/zbud.c
index 3f61304405cb..6348932430b8 100644
--- a/mm/zbud.c
+++ b/mm/zbud.c
@@ -111,10 +111,8 @@ struct zbud_pool {
        struct list_head lru;
        u64 pages_nr;
        const struct zbud_ops *ops;
-#ifdef CONFIG_ZPOOL
        struct zpool *zpool;
        const struct zpool_ops *zpool_ops;
-#endif
 };

 /*
@@ -526,8 +524,6 @@ static u64 zbud_get_pool_size(struct zbud_pool *pool)
  * zpool
  ****************/

-#ifdef CONFIG_ZPOOL
-
 static int zbud_zpool_evict(struct zbud_pool *pool, unsigned long handle)
 {
        if (pool->zpool && pool->zpool_ops && pool->zpool_ops->evict)
@@ -618,7 +614,6 @@ static struct zpool_driver zbud_zpool_driver = {
 };

 MODULE_ALIAS("zpool-zbud");
-#endif /* CONFIG_ZPOOL */

 static int __init init_zbud(void)
 {
@@ -626,19 +621,14 @@ static int __init init_zbud(void)
        BUILD_BUG_ON(sizeof(struct zbud_header) > ZHDR_SIZE_ALIGNED);
        pr_info("loaded\n");

-#ifdef CONFIG_ZPOOL
        zpool_register_driver(&zbud_zpool_driver);
-#endif

        return 0;
 }

 static void __exit exit_zbud(void)
 {
-#ifdef CONFIG_ZPOOL
        zpool_unregister_driver(&zbud_zpool_driver);
-#endif
-
        pr_info("unloaded\n");
 }

> This patch causes several new warnings when CONFIG_ZPOOL is disabled:
> 
> mm/zbud.c:222:26: warning: unused function 'zbud_create_pool' [-Wunused-function]
> mm/zbud.c:246:13: warning: unused function 'zbud_destroy_pool' [-Wunused-function]
> mm/zbud.c:270:12: warning: unused function 'zbud_alloc' [-Wunused-function]
> mm/zbud.c:345:13: warning: unused function 'zbud_free' [-Wunused-function]
> mm/zbud.c:417:12: warning: unused function 'zbud_reclaim_page' [-Wunused-function]
> mm/zbud.c:499:14: warning: unused function 'zbud_map' [-Wunused-function]
> mm/zbud.c:509:13: warning: unused function 'zbud_unmap' [-Wunused-function]
> mm/zbud.c:520:12: warning: unused function 'zbud_get_pool_size' [-Wunused-function]
> 
> It seems to me like all of these functions should be sunk into their
> callers and eliminated entirely as part of this refactoring. I took a
> whack at it but got lost with the kernel docs so someone who is familiar
> with this should probably do it.
> 
> Cheers,
> Nathan
> .
>