From: "Timur Safin" <tsafin@tarantool.org> To: 'Vladislav Shpilevoy' <v.shpilevoy@tarantool.org>, tarantool-patches@dev.tarantool.org, alyapunov@tarantool.org, korablev@tarantool.org Subject: Re: [Tarantool-patches] [PATCH v2 04/10] crc32: align memory access Date: Thu, 28 May 2020 23:11:56 +0300 [thread overview] Message-ID: <048601d6352c$3cdb81a0$b69284e0$@tarantool.org> (raw) In-Reply-To: <c6188bdc4734e30ac313f584e28f56cf2bb07875.1590622225.git.v.shpilevoy@tarantool.org> : -----Original Message----- : From: Vladislav Shpilevoy <v.shpilevoy@tarantool.org> : Subject: [PATCH v2 04/10] crc32: align memory access : : : diff --git a/src/cpu_feature.c b/src/cpu_feature.c : index 98567ccb3..9bf6223de 100644 : --- a/src/cpu_feature.c : +++ b/src/cpu_feature.c : @@ -50,7 +51,7 @@ : : : static uint32_t : -crc32c_hw_byte(uint32_t crc, unsigned char const *data, unsigned int : length) : +crc32c_hw_byte(uint32_t crc, char const *data, unsigned int length) : { : while (length--) { : __asm__ __volatile__( : @@ -68,6 +69,26 @@ crc32c_hw_byte(uint32_t crc, unsigned char const *data, : unsigned int length) : uint32_t : crc32c_hw(uint32_t crc, const char *buf, unsigned int len) : { : + const int align = alignof(unsigned long); : + unsigned long addr = (unsigned long)buf; : + unsigned int not_aligned_prefix = : + ((addr - 1 + align) & ~(align - 1)) - addr; Hmm, hmm... Isn't it simple `addr % align`? Or even `addr & (align - 1)` ? : + /* : + * Calculate CRC32 for the prefix byte-by-byte so as to : + * then use aligned words to calculate the rest. This is : + * twice less loads, because every load takes exactly one : + * word from memory. Not 2 words, which would need to be : + * partially merged then. : + * But the main reason is that unaligned loads are just : + * unsafe, because this is an undefined behaviour. : + */ : + if (not_aligned_prefix < len) { : + crc = crc32c_hw_byte(crc, buf, not_aligned_prefix); : + buf += not_aligned_prefix; : + len -= not_aligned_prefix; : + } else { : + return crc32c_hw_byte(crc, buf, len); : + } : unsigned int iquotient = len / SCALE_F; : unsigned int iremainder = len % SCALE_F; : unsigned long *ptmp = (unsigned long *)buf;
next prev parent reply other threads:[~2020-05-28 20:11 UTC|newest] Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top 2020-05-27 23:32 [Tarantool-patches] [PATCH v2 00/10] Sanitize unaligned access Vladislav Shpilevoy 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 01/10] small: sanitized rlist and new region API Vladislav Shpilevoy 2020-05-28 20:41 ` Timur Safin 2020-05-28 22:56 ` Vladislav Shpilevoy 2020-06-08 23:01 ` Vladislav Shpilevoy 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 10/10] xrow: use unaligned store operation in xrow_to_iovec() Vladislav Shpilevoy 2020-05-28 20:20 ` Timur Safin 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 02/10] cmake: ignore warnings on alignof() and offsetof() Vladislav Shpilevoy 2020-05-28 20:18 ` Timur Safin 2020-05-29 6:24 ` Kirill Yukhin 2020-05-29 22:34 ` Vladislav Shpilevoy 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 03/10] cmake: add option ENABLE_UB_SANITIZER Vladislav Shpilevoy 2020-05-28 20:42 ` Timur Safin 2020-05-29 8:53 ` Sergey Bronnikov 2020-05-29 22:36 ` Vladislav Shpilevoy 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 04/10] crc32: align memory access Vladislav Shpilevoy 2020-05-28 20:11 ` Timur Safin [this message] 2020-05-28 23:23 ` Vladislav Shpilevoy 2020-05-28 23:32 ` Timur Safin 2020-06-08 22:33 ` Vladislav Shpilevoy 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 05/10] sql: make BtCursor's memory aligned Vladislav Shpilevoy 2020-05-28 20:20 ` Timur Safin 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 06/10] region: use aligned allocations where necessary Vladislav Shpilevoy 2020-05-28 20:35 ` Timur Safin 2020-05-28 23:07 ` Vladislav Shpilevoy 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 07/10] vinyl: align statements and bps tree extents Vladislav Shpilevoy 2020-05-28 20:38 ` Timur Safin 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 08/10] tuple: use unaligned store-load for field map Vladislav Shpilevoy 2020-05-28 20:22 ` Timur Safin 2020-05-27 23:32 ` [Tarantool-patches] [PATCH v2 09/10] port: make port_c_entry not PACKED Vladislav Shpilevoy 2020-05-28 20:42 ` Timur Safin 2020-06-03 21:27 ` [Tarantool-patches] [PATCH v2 00/10] Sanitize unaligned access Vladislav Shpilevoy 2020-06-08 22:33 ` Vladislav Shpilevoy
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='048601d6352c$3cdb81a0$b69284e0$@tarantool.org' \ --to=tsafin@tarantool.org \ --cc=alyapunov@tarantool.org \ --cc=korablev@tarantool.org \ --cc=tarantool-patches@dev.tarantool.org \ --cc=v.shpilevoy@tarantool.org \ --subject='Re: [Tarantool-patches] [PATCH v2 04/10] crc32: align memory access' \ /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
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox