| Message ID | 20220210133812.13368-1-bogdan.lezhepekov@suse.com |
|---|---|
| State | Accepted |
| Headers | show |
| Series | [v1] mem/swapping01: Make sure swap is supported by kernel | expand |
Hi! Applied, thanks. I guess that the SAFE_OPEN() caused TBROK on machines without CONFIG_SWAP, right?
Absolutely correct 🙂 Thanks for the review! -Bogdan
On Fri, Feb 11, 2022 at 12:09 AM Cyril Hrubis <chrubis@suse.cz> wrote: > Hi! > Applied, thanks. > > I guess that the SAFE_OPEN() caused TBROK on machines without > CONFIG_SWAP, right? > If so, why not go fix with .needs_config for "CONFIG_SWAP=y" ?
Hi Li > > > On Fri, Feb 11, 2022 at 12:09 AM Cyril Hrubis <chrubis@suse.cz > <mailto:chrubis@suse.cz>> wrote: > > Hi! > Applied, thanks. > > I guess that the SAFE_OPEN() caused TBROK on machines without > CONFIG_SWAP, right? > > > If so, why not go fix with .needs_config for "CONFIG_SWAP=y" ? +1, it is also friendly to docparse. ps: I never listen user complaint for the lack of CONFIG_PORC, so only using .needs_config for "CONFIG_SWAP=y" can work well. Best Regards Yang Xu > > -- > Regards, > Li Wang > > > >
Hi all, I didn't go for CONFIG_SWAP just because wanted to be aligned with syscalls/swapon/swapon03.c and to not multiply patterns. If you see CONFIG_SWAP more beneficial for docparser, I can make another patch. Should I? Cheers, Bogdan
Hi Bogdan > Hi all, > > I didn't go for CONFIG_SWAP just because wanted to be aligned with > syscalls/swapon/swapon03.c and to not multiply patterns. > > If you see CONFIG_SWAP more beneficial for docparser, I can make another > patch. > > Should I? IMO, yes, but let's listen voice from other maintainers. @Petr,@Cyril Best Regards Yang Xu > > Cheers, > Bogdan > ------------------------------------------------------------------------ > *From:* ltp <ltp-bounces+bogdan.lezhepekov=suse.com@lists.linux.it> on > behalf of xuyang2018.jy@fujitsu.com <xuyang2018.jy@fujitsu.com> > *Sent:* Friday, February 11, 2022 3:21 > *To:* Li Wang <liwang@redhat.com> > *Cc:* LTP List <ltp@lists.linux.it> > *Subject:* Re: [LTP] [PATCH v1] mem/swapping01: Make sure swap is > supported by kernel > Hi Li >> >> >> On Fri, Feb 11, 2022 at 12:09 AM Cyril Hrubis <chrubis@suse.cz >> <mailto:chrubis@suse.cz>> wrote: >> >> Hi! >> Applied, thanks. >> >> I guess that the SAFE_OPEN() caused TBROK on machines without >> CONFIG_SWAP, right? >> >> >> If so, why not go fix with .needs_config for "CONFIG_SWAP=y" ? > +1, it is also friendly to docparse. > > ps: I never listen user complaint for the lack of CONFIG_PORC, so only > using .needs_config for "CONFIG_SWAP=y" can work well. > > Best Regards > Yang Xu >> >> -- >> Regards, >> Li Wang >> >> >> >> > > -- > Mailing list info: https://lists.linux.it/listinfo/ltp >
Hi all, > Hi Bogdan > > Hi all, > > I didn't go for CONFIG_SWAP just because wanted to be aligned with > > syscalls/swapon/swapon03.c and to not multiply patterns. > > If you see CONFIG_SWAP more beneficial for docparser, I can make another > > patch. > > Should I? > IMO, yes, but let's listen voice from other maintainers. > @Petr,@Cyril We have 3 options. 1) keep check access("/proc/swaps", F_OK) 2) use .needs_config for "CONFIG_SWAP=y" 3) Add .needs_swap flag to struct tst_test To be honest, I'm not sure, what's the best, probably .needs_config. I worried little bit about .needs_config forcing config to be available in the past, but that should be solved KCONFIG_SKIP_CHECK added by Xu. Kind regards, Petr > Best Regards > Yang Xu > > Cheers, > > Bogdan > > ------------------------------------------------------------------------ > > *From:* ltp <ltp-bounces+bogdan.lezhepekov=suse.com@lists.linux.it> on > > behalf of xuyang2018.jy@fujitsu.com <xuyang2018.jy@fujitsu.com> > > *Sent:* Friday, February 11, 2022 3:21 > > *To:* Li Wang <liwang@redhat.com> > > *Cc:* LTP List <ltp@lists.linux.it> > > *Subject:* Re: [LTP] [PATCH v1] mem/swapping01: Make sure swap is > > supported by kernel > > Hi Li > >> On Fri, Feb 11, 2022 at 12:09 AM Cyril Hrubis <chrubis@suse.cz > >> <mailto:chrubis@suse.cz>> wrote: > >> Hi! > >> Applied, thanks. > >> I guess that the SAFE_OPEN() caused TBROK on machines without > >> CONFIG_SWAP, right? > >> If so, why not go fix with .needs_config for "CONFIG_SWAP=y" ? > > +1, it is also friendly to docparse. > > ps: I never listen user complaint for the lack of CONFIG_PORC, so only > > using .needs_config for "CONFIG_SWAP=y" can work well. > > Best Regards > > Yang Xu > >> -- > >> Regards, > >> Li Wang
Hi! > 1) keep check access("/proc/swaps", F_OK) > 2) use .needs_config for "CONFIG_SWAP=y" > 3) Add .needs_swap flag to struct tst_test > > To be honest, I'm not sure, what's the best, probably .needs_config. > I worried little bit about .needs_config forcing config to be available in the > past, but that should be solved KCONFIG_SKIP_CHECK added by Xu. I guess that the easiest solution would be .needs_kconfig, we ended up depending on it in quite a few testcases anyway so I would say that nobody is runing LTP without having kernel config exproted anyways.
Hi, Ok, I will send out a new patch. Do I need to update syscalls/swapon/swapon03.c accordingly to make it a pattern? Cheers, Bogdan
On Fri, Feb 11, 2022 at 9:16 PM Bogdan Lezhepekov < bogdan.lezhepekov@suse.com> wrote: > Hi, > > Ok, I will send out a new patch. > > Do I need to update syscalls/swapon/swapon03.c accordingly to make it a > pattern? > Yes, that would be great to have. Thanks~
diff --git a/testcases/kernel/mem/swapping/swapping01.c b/testcases/kernel/mem/swapping/swapping01.c index 1a494bf69..3b4f0a61a 100644 --- a/testcases/kernel/mem/swapping/swapping01.c +++ b/testcases/kernel/mem/swapping/swapping01.c @@ -155,11 +155,18 @@ static void check_swapping(void) SAFE_WAITPID(pid, &status, 0); } +static void setup(void) +{ + if (access("/proc/swaps", F_OK)) + tst_brk(TCONF, "swap not supported by kernel"); +} + static struct tst_test test = { .needs_root = 1, .forks_child = 1, .min_mem_avail = 10, .test_all = test_swapping, + .setup = setup, .tags = (const struct tst_tag[]) { {"linux-git", "50a15981a1fa"}, {}
Previous version caused false positive result on machines support no swap in kernel. Signed-off-by: Bogdan Lezhepekov <bogdan.lezhepekov@suse.com> --- testcases/kernel/mem/swapping/swapping01.c | 7 +++++++ 1 file changed, 7 insertions(+)