From: Jiri Slaby <jslaby@suse.cz>
To: gregkh@linuxfoundation.org
Cc: linux-serial@vger.kernel.org, linux-kernel@vger.kernel.org,
Jiri Slaby <jslaby@suse.cz>
Subject: [PATCH 09/16] vc_screen: eliminate ifdefs from vcs_write_buf
Date: Tue, 18 Aug 2020 10:56:59 +0200 [thread overview]
Message-ID: <20200818085706.12163-9-jslaby@suse.cz> (raw)
In-Reply-To: <20200818085706.12163-1-jslaby@suse.cz>
Introduce a new inline function called vc_compile_le16 and do the shifts
and ORs there. Depending on LE x BE.
I tried cpu_to_le16, but it ends up with worse assembly on BE for
whatever reason -- the compiler seems to be unable to optimize the swap.
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
---
drivers/tty/vt/vc_screen.c | 29 +++++++++++++++++------------
1 file changed, 17 insertions(+), 12 deletions(-)
diff --git a/drivers/tty/vt/vc_screen.c b/drivers/tty/vt/vc_screen.c
index 10a26fd5f1b7..36b967825f68 100644
--- a/drivers/tty/vt/vc_screen.c
+++ b/drivers/tty/vt/vc_screen.c
@@ -481,6 +481,19 @@ static u16 *vcs_write_buf_noattr(struct vc_data *vc, const char *con_buf,
return org;
}
+/*
+ * Compilers (gcc 10) are unable to optimize the swap in cpu_to_le16. So do it
+ * the poor man way.
+ */
+static inline u16 vc_compile_le16(u8 hi, u8 lo)
+{
+#ifdef __BIG_ENDIAN
+ return (lo << 8u) | hi;
+#else
+ return (hi << 8u) | lo;
+#endif
+}
+
static u16 *vcs_write_buf(struct vc_data *vc, const char *con_buf,
unsigned int pos, unsigned int count, bool viewed, u16 **org0)
{
@@ -513,13 +526,8 @@ static u16 *vcs_write_buf(struct vc_data *vc, const char *con_buf,
if (pos & 1) {
count--;
c = *con_buf++;
-#ifdef __BIG_ENDIAN
- vcs_scr_writew(vc, c |
- (vcs_scr_readw(vc, org) & 0xff00), org);
-#else
- vcs_scr_writew(vc, (c << 8) |
- (vcs_scr_readw(vc, org) & 0xff), org);
-#endif
+ vcs_scr_writew(vc, vc_compile_le16(c, vcs_scr_readw(vc, org)),
+ org);
org++;
pos++;
if (++col == maxcol) {
@@ -551,11 +559,8 @@ static u16 *vcs_write_buf(struct vc_data *vc, const char *con_buf,
/* odd pos -- the remaining character */
c = *con_buf++;
-#ifdef __BIG_ENDIAN
- vcs_scr_writew(vc, (vcs_scr_readw(vc, org) & 0xff) | (c << 8), org);
-#else
- vcs_scr_writew(vc, (vcs_scr_readw(vc, org) & 0xff00) | c, org);
-#endif
+ vcs_scr_writew(vc, vc_compile_le16(vcs_scr_readw(vc, org) >> 8, c),
+ org);
return org;
}
--
2.28.0
next prev parent reply other threads:[~2020-08-18 8:57 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-18 8:56 [PATCH 01/16] vt: make vc_data pointers const in selection.h Jiri Slaby
2020-08-18 8:56 ` [PATCH 02/16] vt: declare xy for get/putconsxy properly Jiri Slaby
2020-08-18 11:43 ` Greg KH
2020-08-19 7:09 ` Jiri Slaby
2020-08-18 8:56 ` [PATCH 03/16] vc: propagate "viewed as bool" from screenpos up Jiri Slaby
2020-08-18 8:56 ` [PATCH 04/16] vc_screen: document and cleanup vcs_vc Jiri Slaby
2020-08-18 8:56 ` [PATCH 05/16] vc_screen: rewrite vcs_size to accept vc, not inode Jiri Slaby
2020-08-18 8:56 ` [PATCH 06/16] vc_screen: sanitize types in vcs_write Jiri Slaby
2020-08-18 8:56 ` [PATCH 07/16] vc_screen: extract vcs_write_buf_noattr Jiri Slaby
2020-08-18 8:56 ` [PATCH 08/16] vc_screen: extract vcs_write_buf Jiri Slaby
2020-08-18 8:56 ` Jiri Slaby [this message]
2020-08-18 8:57 ` [PATCH 10/16] vc_screen: sanitize types in vcs_read Jiri Slaby
2020-08-18 8:57 ` [PATCH 11/16] vs_screen: kill tmp_count from vcs_read Jiri Slaby
2020-08-18 8:57 ` [PATCH 12/16] vc_screen: extract vcs_read_buf_uni Jiri Slaby
2020-08-18 8:57 ` [PATCH 13/16] vc_screen: extract vcs_read_buf_noattr Jiri Slaby
2020-08-18 8:57 ` [PATCH 14/16] vc_screen: extract vcs_read_buf Jiri Slaby
2020-08-18 8:57 ` [PATCH 15/16] vc_screen: extract vcs_read_buf_header Jiri Slaby
2020-08-25 16:48 ` Peilin Ye
2020-08-25 16:59 ` Greg KH
2020-08-25 17:12 ` Peilin Ye
2020-08-18 8:57 ` [PATCH 16/16] vc_screen: prune macros Jiri Slaby
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=20200818085706.12163-9-jslaby@suse.cz \
--to=jslaby@suse.cz \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-serial@vger.kernel.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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 NNTP newsgroup(s).