* [PATCH 1/2] scripts: imx-image: Do not pad image
2015-07-16 7:48 [PATCH v2]: imx-image changes Sascha Hauer
@ 2015-07-16 7:48 ` Sascha Hauer
2015-07-16 7:48 ` [PATCH 2/2] scripts: imx-image: Make in-place capable Sascha Hauer
1 sibling, 0 replies; 3+ messages in thread
From: Sascha Hauer @ 2015-07-16 7:48 UTC (permalink / raw)
To: Barebox List
We have to pad the load size to the next 4k boundary, but only for the
HAB4 case we actually care what data is loaded in the rest of the image.
This lets the padding depend on the prepare_sign option.
Background for this patch is a new yet-to-be-introduced image loading
mechanism for i.MX. This new mechanism only loads the PBL portion of
the image to memory, and not the whole image anymore. This means that
the image in this case changes from:
i.MX header (with loadsize: whole image), PBL, payload, padding
to:
i.MX header (with loadsize: header + PBL + padding), PBL, padding, payload
With the padding between PBL and payload we are no longer able to find
the payload, so we cannot add the padding there.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
scripts/imx/imx-image.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/scripts/imx/imx-image.c b/scripts/imx/imx-image.c
index e765c1d..f0e8ca3 100644
--- a/scripts/imx/imx-image.c
+++ b/scripts/imx/imx-image.c
@@ -825,7 +825,7 @@ int main(int argc, char *argv[])
/* pad until next 4k boundary */
now = 4096 - now;
- if (now) {
+ if (prepare_sign && now) {
memset(buf, 0x5a, now);
ret = xwrite(outfd, buf, now);
--
2.1.4
_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox
^ permalink raw reply [flat|nested] 3+ messages in thread
* [PATCH 2/2] scripts: imx-image: Make in-place capable
2015-07-16 7:48 [PATCH v2]: imx-image changes Sascha Hauer
2015-07-16 7:48 ` [PATCH 1/2] scripts: imx-image: Do not pad image Sascha Hauer
@ 2015-07-16 7:48 ` Sascha Hauer
1 sibling, 0 replies; 3+ messages in thread
From: Sascha Hauer @ 2015-07-16 7:48 UTC (permalink / raw)
To: Barebox List
Read in the source image completely before starting to write the output
image. This makes it possible to pass the same file as input and output.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
scripts/imx/imx-image.c | 47 ++++++++++++++++++++++++-----------------------
1 file changed, 24 insertions(+), 23 deletions(-)
diff --git a/scripts/imx/imx-image.c b/scripts/imx/imx-image.c
index f0e8ca3..e8d9dbf 100644
--- a/scripts/imx/imx-image.c
+++ b/scripts/imx/imx-image.c
@@ -666,7 +666,8 @@ int main(int argc, char *argv[])
char *imagename = NULL;
char *outfile = NULL;
void *buf;
- size_t image_size = 0, load_size;
+ size_t image_size = 0, load_size, insize;
+ void *infile;
struct stat s;
int infd, outfd;
int dcd_only = 0;
@@ -779,6 +780,24 @@ int main(int argc, char *argv[])
exit(1);
}
+ infd = open(imagename, O_RDONLY);
+ if (infd < 0) {
+ perror("open");
+ exit(1);
+ }
+
+ ret = fstat(infd, &s);
+ if (ret)
+ return ret;
+
+ insize = s.st_size;
+ infile = malloc(insize);
+ if (!infile)
+ exit(1);
+
+ xread(infd, infile, insize);
+ close(infd);
+
outfd = open(outfile, O_WRONLY | O_CREAT | O_TRUNC, S_IRUSR | S_IWUSR);
if (outfd < 0) {
perror("open");
@@ -799,32 +818,14 @@ int main(int argc, char *argv[])
}
}
- infd = open(imagename, O_RDONLY);
- if (infd < 0) {
- perror("open");
+ ret = xwrite(outfd, infile, insize);
+ if (ret) {
+ perror("write");
exit(1);
}
- while (image_size) {
- now = image_size < 4096 ? image_size : 4096;
-
- ret = xread(infd, buf, now);
- if (ret) {
- perror("read");
- exit(1);
- }
-
- ret = xwrite(outfd, buf, now);
- if (ret) {
- perror("write");
- exit(1);
- }
-
- image_size -= now;
- }
-
/* pad until next 4k boundary */
- now = 4096 - now;
+ now = 4096 - (insize % 4096);
if (prepare_sign && now) {
memset(buf, 0x5a, now);
--
2.1.4
_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox
^ permalink raw reply [flat|nested] 3+ messages in thread