From mboxrd@z Thu Jan 1 00:00:00 1970 From: Guenter Roeck Subject: Re: [PATCH v5 6/9] bitops: let optimize out non-atomic bitops on compile-time constants Date: Fri, 15 Jul 2022 06:49:46 -0700 Message-ID: <8c949bd4-d25a-d5f5-49be-59d52e4b6c9d@roeck-us.net> References: <20220624121313.2382500-1-alexandr.lobakin@intel.com> <20220624121313.2382500-7-alexandr.lobakin@intel.com> <20220715000402.GA512558@roeck-us.net> <20220715132633.61480-1-alexandr.lobakin@intel.com> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=sender:message-id:date:mime-version:user-agent:subject :content-language:to:cc:references:from:in-reply-to :content-transfer-encoding; bh=Rs46+iydynCKHLHZ1GP3rF0W0E38SP1veulXa6gwRZ4=; b=U2LPKLM/AjT1a2LjEeJbIoC7cMGOq1yROmt0cZz2Ais3HJJbyhb17PcE+5FV+ybUP8 +DSpyzNvjpXy8F2wrmZ0lTlqucMS7kjrw4Q9h/Tne1y2K9xW3tm2zMzEVInk6+Y2kNSK xK5+NjqnN7jZRM20X7NKghDHD1GEMUwYqciGlk2cObJdUKWdRpmYfuyS/zSPa7rtAs0J /eL2zWxi2I+kPcM56S+S1qVAVJaJURtunnq0/WI/1FWTaavZX0kfTCaS9YaUXyinPP0X /r6r+fcRBASD1bIPLyE3iFMUJV6F7KavxvqQHCWQ7TydHLHLIKdC54p0qZnsEtBjn01F 8XWg== Sender: Guenter Roeck Content-Language: en-US In-Reply-To: <20220715132633.61480-1-alexandr.lobakin@intel.com> List-ID: Content-Type: text/plain; charset="us-ascii"; format="flowed" To: Alexander Lobakin Cc: Arnd Bergmann , Yury Norov , Andy Shevchenko , Mark Rutland , Matt Turner , Brian Cain , Geert Uytterhoeven , Yoshinori Sato , Rich Felker , "David S. Miller" , Kees Cook , "Peter Zijlstra (Intel)" , Marco Elver , Borislav Petkov , Tony Luck , Maciej Fijalkowski , Jesse Brandeburg , Greg Kroah-Hartman , Nathan Chancellor , Nick On 7/15/22 06:26, Alexander Lobakin wrote: > From: Guenter Roeck > Date: Thu, 14 Jul 2022 17:04:02 -0700 > >> On Fri, Jun 24, 2022 at 02:13:10PM +0200, Alexander Lobakin wrote: >>> Currently, many architecture-specific non-atomic bitop >>> implementations use inline asm or other hacks which are faster or > > [...] > >>> Cc: Mark Rutland >>> Signed-off-by: Alexander Lobakin >>> Reviewed-by: Marco Elver >> >> Building i386:allyesconfig ... failed >> -------------- >> Error log: >> arch/x86/platform/olpc/olpc-xo1-sci.c: In function 'send_ebook_state': >> arch/x86/platform/olpc/olpc-xo1-sci.c:83:63: error: logical not is only applied to the left hand side of comparison > > Looks like a trigger, not a cause... Anyway, this construct: > > unsigned char state; > > [...] > > if (!!test_bit(SW_TABLET_MODE, ebook_switch_idev->sw) == state) > > doesn't look legit enough. > That redundant double-negation [of boolean value], together with > comparing boolean to char, provokes compilers to think the author > made logical mistakes here, although it works as expected. > Could you please try (if it's not automated build which you can't > modify) the following: > Agreed, the existing code seems wrong. The change below looks correct and fixes the problem. Feel free to add Reviewed-and-tested-by: Guenter Roeck to the real patch. Thanks, Guenter > --- a/arch/x86/platform/olpc/olpc-xo1-sci.c > +++ b/arch/x86/platform/olpc/olpc-xo1-sci.c > @@ -80,7 +80,7 @@ static void send_ebook_state(void) > return; > } > > - if (!!test_bit(SW_TABLET_MODE, ebook_switch_idev->sw) == state) > + if (test_bit(SW_TABLET_MODE, ebook_switch_idev->sw) == !!state) > return; /* Nothing new to report. */ > > input_report_switch(ebook_switch_idev, SW_TABLET_MODE, state); > --- > > We'd take it into the bitmap tree then. The series revealed > a fistful of existing code issues already :) > >> >> Bisect log attached. >> >> Guenter >> >> --- >> # bad: [4662b7adea50bb62e993a67f611f3be625d3df0d] Add linux-next specific files for 20220713 >> # good: [32346491ddf24599decca06190ebca03ff9de7f8] Linux 5.19-rc6 >> git bisect start 'HEAD' 'v5.19-rc6' >> # good: [8b7e002d8bc6e17c94092d25e7261db4e6e5f2cc] Merge branch 'drm-next' of git://git.freedesktop.org/git/drm/drm.git >> git bisect good 8b7e002d8bc6e17c94092d25e7261db4e6e5f2cc >> # good: [07f6d21d6e33c1e28e24ae84e9d26e4e7d4853f5] Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/kvms390/linux.git >> git bisect good 07f6d21d6e33c1e28e24ae84e9d26e4e7d4853f5 >> # good: [5ff085e5d4f6700e03635d5e700f52163a6dc2a7] Merge branch 'staging-next' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging.git >> git bisect good 5ff085e5d4f6700e03635d5e700f52163a6dc2a7 >> # good: [eb9e3fdbdd8b61ef0f4bee23259fe6ab69e463ab] Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/crng/random.git >> git bisect good eb9e3fdbdd8b61ef0f4bee23259fe6ab69e463ab >> # good: [9f2183cd961e5ddb7954eafb6bb01a495c6a9c7b] hexagon/mm: enable ARCH_HAS_VM_GET_PAGE_PROT >> git bisect good 9f2183cd961e5ddb7954eafb6bb01a495c6a9c7b >> # bad: [e878aa5faf9ac8c0b5d0c3f293389c194c250fff] Merge branch 'mm-nonmm-stable' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm >> git bisect bad e878aa5faf9ac8c0b5d0c3f293389c194c250fff >> # good: [cf95d50205f62c4f5f538676def847292cf39fa9] fs: don't call ->writepage from __mpage_writepage >> git bisect good cf95d50205f62c4f5f538676def847292cf39fa9 >> # good: [5103cbfd92d3587713476f94f9485b96e02f0146] Merge branch 'for-next/execve' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux.git >> git bisect good 5103cbfd92d3587713476f94f9485b96e02f0146 >> # good: [ee56c3e8eec166f4e4a2ca842b7804d14f3a0208] Merge branch 'master' into mm-nonmm-stable >> git bisect good ee56c3e8eec166f4e4a2ca842b7804d14f3a0208 >> # bad: [dc34d5036692c614eef23c1130ee42a201c316bf] lib: test_bitmap: add compile-time optimization/evaluations assertions >> git bisect bad dc34d5036692c614eef23c1130ee42a201c316bf >> # good: [bb7379bfa680bd48b468e856475778db2ad866c1] bitops: define const_*() versions of the non-atomics >> git bisect good bb7379bfa680bd48b468e856475778db2ad866c1 >> # bad: [b03fc1173c0c2bb8fad61902a862985cecdc4b1b] bitops: let optimize out non-atomic bitops on compile-time constants >> git bisect bad b03fc1173c0c2bb8fad61902a862985cecdc4b1b >> # good: [e69eb9c460f128b71c6b995d75a05244e4b6cc3e] bitops: wrap non-atomic bitops with a transparent macro >> git bisect good e69eb9c460f128b71c6b995d75a05244e4b6cc3e >> # first bad commit: [b03fc1173c0c2bb8fad61902a862985cecdc4b1b] bitops: let optimize out non-atomic bitops on compile-time constants > > Thanks, > Olek