On 10/26/22 01:35, Simon Glass wrote:
Hi Heinrich,
On Sat, 22 Oct 2022 at 03:21, Heinrich Schuchardt
<heinrich.schucha...@canonical.com> wrote:
We may enter the command line interface in a state where on the remote
console the cursor is not shown. Send an escape sequence to enable it.
Signed-off-by: Heinrich Schuchardt <heinrich.schucha...@canonical.com>
---
common/main.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/common/main.c b/common/main.c
index 682f3359ea..4e7e7b17d6 100644
--- a/common/main.c
+++ b/common/main.c
@@ -7,6 +7,7 @@
/* #define DEBUG */
#include <common.h>
+#include <ansi.h>
#include <autoboot.h>
#include <bootstage.h>
#include <cli.h>
@@ -66,6 +67,9 @@ void main_loop(void)
autoboot_command(s);
+ if (!CONFIG_IS_ENABLED(DM_VIDEO) || CONFIG_IS_ENABLED(VIDEO_ANSI))
+ printf(ANSI_CURSOR_SHOW "\n");
Could we create a library which emits these codes? Like ansi_cursor_show() ?
The question is not if we could but if we should.
Up to now we tried to call printf() only once where ANSI_* is only one
part of the output string.
E.g. cmd/eficonfig.c:415:
printf(ANSI_CLEAR_CONSOLE
ANSI_CURSOR_POSITION
ANSI_CURSOR_SHOW, 1, 1);
This minimizes the number of call library calls but may not be very elegant.
Creating a library function for ANSI_CURSOR_SHOW alone does not make
much sense to me. Should you want to convert our code base from using
ANSI_* to library functions this should cover the whole code base.
So I think we should let this patch pass as it solves a current problem
and leave creating a ANSI_* function library to a separate exercise.
Best regards
Heinrich
+
cli_loop();
panic("No CLI available");
}
--
2.37.2
Regards,
Simon
