Git for Windows crash when I try to use git difftool -d on a file added with -N option #2677

sachinrk · 2020-06-09T18:03:28Z

I was not able to find an open or closed issue matching what I'm seeing

Setup

Which version of Git for Windows are you using? Is it 32-bit or 64-bit?
git version 2.27.0.windows.1
64 bit.

I recently upgraded the git version, but the issue was reproducible on older versions as well.

$ git --version --build-options

git version 2.27.0.windows.1
cpu: x86_64
built from commit: 907ab1011dce9112700498e034b974ba60f8b407
sizeof-long: 4
sizeof-size_t: 8

Which version of Windows are you running? Vista, 7, 8, 10? Is it 32-bit or 64-bit?

$ cmd.exe /c ver

Microsoft Windows [Version 10.0.18363.836

What options did you set as part of the installation? Or did you choose the
defaults?

# One of the following:
> type "C:\Program Files\Git\etc\install-options.txt"
> type "C:\Program Files (x86)\Git\etc\install-options.txt"
> type "%USERPROFILE%\AppData\Local\Programs\Git\etc\install-options.txt"
$ cat /etc/install-options.txt

type "C:\Program Files\Git\etc\install-options.txt"

Editor Option: VIM
Custom Editor Path:
Path Option: Cmd
SSH Option: OpenSSH
Tortoise Option: false
CURL Option: OpenSSL
CRLF Option: CRLFAlways
Bash Terminal Option: MinTTY
Git Pull Behavior Option: Merge
Performance Tweaks FSCache: Enabled
Use Credential Manager: Enabled
Enable Symlinks: Disabled
Enable Pseudo Console Support: Disabled

Any other interesting things about your environment that might be related
to the issue you're seeing?

Nothing interesting.

Details

Which terminal/shell are you running Git from? e.g Bash/CMD/PowerShell/other

CMD

What commands did you run to trigger this issue? If you can provide a
Minimal, Complete, and Verifiable example
this will help us understand the issue.

In a new branch, simply 
notepad++ a.txt
git add -N a.txt
git difftool -d

What did you expect to occur after running these commands?

No crash, and that git would open my difftool with the diff of the file

What actually happened instead?

d:\RoutingPlane\sources\dev\CAFE\NanoProxy>git difftool -d
error: unable to read sha1 file of C:\Users\sakul\AppData\Local\Temp\git-difftool.a32148/right/build/configuration/public/xml/a.txt (4b825dc642cb6eb9a060e54bf8d69288fbee4904)
error: could not write 'build/configuration/public/xml/a.txt'

Crash dialog saying Git for Windows has stopped working

If the problem was occurring with a specific repository, can you provide the
URL to that repository to help us with testing?

Not specific to the repository

The text was updated successfully, but these errors were encountered:

dscho · 2020-06-09T22:03:25Z

I can reproduce, even on Linux (read: this problem is not Windows-specific):

diff --git a/t/t7800-difftool.sh b/t/t7800-difftool.sh
index e66af3682f8..731ed73f421 100755
--- a/t/t7800-difftool.sh
+++ b/t/t7800-difftool.sh
@@ -720,6 +720,14 @@ test_expect_success SYMLINKS 'difftool --dir-diff handles modified symlinks' '
 	test_cmp expect actual
 '
 
+test_expect_success 'add -N and difftool -d' '
+	test_when_finished git reset --hard &&
+
+	test_write_lines A B C >intent-to-add &&
+	git add -N intent-to-add &&
+	git difftool --dir-diff --extcmd ls
+'
+
 test_expect_success 'outside worktree' '
 	echo 1 >1 &&
 	echo 2 >2 &&

And this seems to fix it:

diff --git a/diff.c b/diff.c
index d1ad6a3c4ad..3503dcdf0db 100644
--- a/diff.c
+++ b/diff.c
@@ -5684,6 +5684,9 @@ static void diff_flush_raw(struct diff_filepair *p, struct diff_options *opt)
 	int line_termination = opt->line_termination;
 	int inter_name_termination = line_termination ? '\t' : '\0';
 
+	diff_fill_oid_info(p->one, opt->repo->index);
+	diff_fill_oid_info(p->two, opt->repo->index);
+
 	fprintf(opt->file, "%s", diff_line_prefix(opt));
 	if (!(opt->output_format & DIFF_FORMAT_NAME_STATUS)) {
 		fprintf(opt->file, ":%06o %06o %s ", p->one->mode, p->two->mode,

sachinrk · 2020-06-09T22:24:50Z

I can reproduce, even on Linux (read: this problem is not Windows-specific):

diff --git a/t/t7800-difftool.sh b/t/t7800-difftool.sh
index e66af3682f8..731ed73f421 100755
--- a/t/t7800-difftool.sh
+++ b/t/t7800-difftool.sh
@@ -720,6 +720,14 @@ test_expect_success SYMLINKS 'difftool --dir-diff handles modified symlinks' '
 	test_cmp expect actual
 '
 
+test_expect_success 'add -N and difftool -d' '
+	test_when_finished git reset --hard &&
+
+	test_write_lines A B C >intent-to-add &&
+	git add -N intent-to-add &&
+	git difftool --dir-diff --extcmd ls
+'
+
 test_expect_success 'outside worktree' '
 	echo 1 >1 &&
 	echo 2 >2 &&

And this seems to fix it:

diff --git a/diff.c b/diff.c
index d1ad6a3c4ad..3503dcdf0db 100644
--- a/diff.c
+++ b/diff.c
@@ -5684,6 +5684,9 @@ static void diff_flush_raw(struct diff_filepair *p, struct diff_options *opt)
 	int line_termination = opt->line_termination;
 	int inter_name_termination = line_termination ? '\t' : '\0';
 
+	diff_fill_oid_info(p->one, opt->repo->index);
+	diff_fill_oid_info(p->two, opt->repo->index);
+
 	fprintf(opt->file, "%s", diff_line_prefix(opt));
 	if (!(opt->output_format & DIFF_FORMAT_NAME_STATUS)) {
 		fprintf(opt->file, ":%06o %06o %s ", p->one->mode, p->two->mode,

Thanks for the quick fix! :)

dscho · 2020-06-10T09:22:58Z

It's not actually a fix because it fails the test suite, pointing out a rather big conceptual problem (I did not figure out how to distinguish between a cache_entry that is --intent-to-add vs one that is simply in need of refreshing).

In `run_diff_files()`, files that have been staged with the intention to add are queued without a valid OID in the `diff_filepair`. When the output mode is, say, `DIFF_FORMAT_PATCH`, the `diff_fill_oid_info()` function, called from `run_diff()`, will remedy that situation by reading the file contents from disk. However, when the output mode is `DIFF_FORMAT_RAW`, that does not hold true, and the output will contain a bogus OID (and the flag `M` for "modified" instead of the correct `A` for "added"). As a consequence, `git difftool -d` (which relies on `git diff-files --raw`'s output) does not work correctly. Let's fix this specifically by imitating `diff_fill_oid_info()`. Note: we can only do that for diff formats that do not actually need the file contents, such as `DIFF_FORMAT_PATCH`: `run_diff()` would try to read the blob contents, but that blob might not have been written to Git's object database. This fixes git-for-windows#2677 Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed. Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In `run_diff_files()`, files that have been staged with the intention to add are queued without a valid OID in the `diff_filepair`. When the output mode is, say, `DIFF_FORMAT_PATCH`, the `diff_fill_oid_info()` function, called from `run_diff()`, will remedy that situation by reading the file contents from disk. However, when the output mode is `DIFF_FORMAT_RAW`, that does not hold true, and the output will contain a bogus OID (and the flag `M` for "modified" instead of the correct `A` for "added"). As a consequence, `git difftool -d` (which relies on `git diff-files --raw`'s output) does not work correctly. Let's fix this specifically by imitating `diff_fill_oid_info()`. Note: we can only do that for diff formats that do not actually need the file contents, such as `DIFF_FORMAT_PATCH`: `run_diff()` would try to read the blob contents, but that blob might not have been written to Git's object database. This fixes git-for-windows#2677 This patch _also_ fixes the expectations set by the regression test introduced in feea694 (diff-files: treat "i-t-a" files as "not-in-index", 2020-06-20). Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed. Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In `run_diff_files()`, files that have been staged with the intention to add are queued without a valid OID in the `diff_filepair`. When the output mode is, say, `DIFF_FORMAT_PATCH`, the `diff_fill_oid_info()` function, called from `run_diff()`, will remedy that situation by reading the file contents from disk. However, when the output mode is `DIFF_FORMAT_RAW`, that does not hold true, and the output will contain a bogus OID (and the flag `M` for "modified" instead of the correct `A` for "added"). As a consequence, `git difftool -d` (which relies on `git diff-files --raw`'s output) does not work correctly. Let's fix this specifically by imitating `diff_fill_oid_info()`. Note: we can only do that for diff formats that do not actually need the file contents, such as `DIFF_FORMAT_PATCH`: `run_diff()` would try to read the blob contents, but that blob might not have been written to Git's object database. This fixes git-for-windows#2677 This patch _also_ fixes the expectations set by the regression test introduced in feea694 (diff-files: treat "i-t-a" files as "not-in-index", 2020-06-20). Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed. Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>

In `run_diff_files()`, files that have been staged with the intention to add are queued without a valid OID in the `diff_filepair`. When the output mode is, say, `DIFF_FORMAT_PATCH`, the `diff_fill_oid_info()` function, called from `run_diff()`, will remedy that situation by reading the file contents from disk. However, when the output mode is `DIFF_FORMAT_RAW`, that does not hold true, and the output will contain a bogus OID (and the flag `M` for "modified" instead of the correct `A` for "added"). As a consequence, `git difftool -d` (which relies on `git diff-files --raw`'s output) does not work correctly. Let's fix this specifically by imitating `diff_fill_oid_info()`. Note: we can only do that for diff formats that do not actually need the file contents, such as `DIFF_FORMAT_PATCH`: `run_diff()` would try to read the blob contents, but that blob might not have been written to Git's object database. This fixes git-for-windows#2677 This patch _also_ fixes the expectations set by the regression test introduced in feea694 (diff-files: treat "i-t-a" files as "not-in-index", 2020-06-20). Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed. Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed. Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed: it reported intent-to-add files with the empty _tree_ as the post-image OID, when we need to show an all-zero (or, "null") OID instead, to indicate to the caller that they have to look at the worktree file. The symptom of that problem shown by `git difftool` was this: error: unable to read sha1 file of <path> (<empty-tree-OID>) error: could not write '<filename>' Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed: it reported intent-to-add files with the empty _tree_ as the post-image OID, when we need to show an all-zero (or, "null") OID instead, to indicate to the caller that they have to look at the worktree file. The symptom of that problem shown by `git difftool` was this: error: unable to read sha1 file of <path> (<empty-tree-OID>) error: could not write '<filename>' Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>

dscho · 2020-07-02T14:58:44Z

Just a quick update that gitgitgadget#654 made it into the next branch and will hopefully be included in Git v2.28.0. If not, my plan is to include it in Git for Windows v2.28.0, anyway.

sachinrk · 2020-07-02T16:14:06Z

Sounds great. Thank you for the update.

…

-Sachin.

On Thu, Jul 2, 2020, 7:58 AM Johannes Schindelin ***@***.***> wrote: Just a quick update that gitgitgadget#654 <gitgitgadget#654> made it into the next branch and will hopefully be included in Git v2.28.0. If not, my plan is to include it in Git *for Windows* v2.28.0, anyway. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#2677 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AABGY47KBMSM7SHBXMECK6TRZSODHANCNFSM4NZUPQ7A> .

In git-for-windows#2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed: it reported intent-to-add files with the empty _tree_ as the post-image OID, when we need to show an all-zero (or, "null") OID instead, to indicate to the caller that they have to look at the worktree file. The symptom of that problem shown by `git difftool` was this: error: unable to read sha1 file of <path> (<empty-tree-OID>) error: could not write '<filename>' Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

In #2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed: it reported intent-to-add files with the empty _tree_ as the post-image OID, when we need to show an all-zero (or, "null") OID instead, to indicate to the caller that they have to look at the worktree file. The symptom of that problem shown by `git difftool` was this: error: unable to read sha1 file of <path> (<empty-tree-OID>) error: could not write '<filename>' Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

dscho · 2020-07-08T13:42:09Z

@sachinrk would you mind testing the latest snapshot?

sachinrk · 2020-07-08T14:04:02Z

Won't mind at all. What do I have to do?

…

On Wed, Jul 8, 2020, 6:42 AM Johannes Schindelin ***@***.***> wrote: @sachinrk <https://github.com/sachinrk> would you mind testing the latest snapshot <https://wingit.blob.core.windows.net/files/index.html>? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#2677 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AABGY4376FKZKQQPD6AKI53R2RZUDANCNFSM4NZUPQ7A> .

dscho · 2020-07-08T14:15:35Z

Won't mind at all. What do I have to do?

Just install the latest snapshot and re-test whether the bug you reported is fixed or not ;-)

sachinrk · 2020-07-08T23:33:30Z

Can confirm the issue is resolved with the latest git for windows snapshot

d:\RoutingPlane\sources\test\CAFE\NanoProxyTests>git --version
git version 2.27.0.windows.1.18.g51f9af323b.20200707200701

d:\RoutingPlane\sources\test\CAFE\NanoProxyTests>notepad++ a.txt

d:\RoutingPlane\sources\test\CAFE\NanoProxyTests>git st
On branch user/sakul/entity_body_support
Your branch is ahead of 'origin/user/sakul/entity_body_support' by 1 commit.
(use "git push" to publish your local commits)

Untracked files:
(use "git add ..." to include in what will be committed)
a.txt

nothing added to commit but untracked files present (use "git add" to track)

d:\RoutingPlane\sources\test\CAFE\NanoProxyTests>git add -N a.txt

d:\RoutingPlane\sources\test\CAFE\NanoProxyTests>git st
On branch user/sakul/entity_body_support
Your branch is ahead of 'origin/user/sakul/entity_body_support' by 1 commit.
(use "git push" to publish your local commits)

Changes not staged for commit:
(use "git add ..." to update what will be committed)
(use "git restore ..." to discard changes in working directory)
new file: a.txt

no changes added to commit (use "git add" and/or "git commit -a")

d:\RoutingPlane\sources\test\CAFE\NanoProxyTests>git difftool -d

d:\RoutingPlane\sources\test\CAFE\NanoProxyTests>

In #2677, a `git difftool -d` problem was reported. The underlying cause was a bug in `git diff-files --raw` that we just fixed: it reported intent-to-add files with the empty _tree_ as the post-image OID, when we need to show an all-zero (or, "null") OID instead, to indicate to the caller that they have to look at the worktree file. The symptom of that problem shown by `git difftool` was this: error: unable to read sha1 file of <path> (<empty-tree-OID>) error: could not write '<filename>' Make sure that the reported `difftool` problem stays fixed. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

sachinrk · 2020-07-20T18:38:35Z

Is it possible to port this fix to git used with vfs? specifically - 2.23.0.vfs.1.1 and release an update?

dscho · 2020-08-10T14:28:47Z

@sachinrk I believe that the latest VFS release comes with this fix.

dscho mentioned this issue Jun 9, 2020

Fix difftool problem with intent-to-add files gitgitgadget/git#654

Closed

rimrul added the git-upstream label Jun 24, 2020

dscho mentioned this issue Jul 3, 2020

Fix git difftool with intent-to-add files #2733

Merged

dscho added this to the Next release milestone Jul 3, 2020

dscho closed this as completed in #2733 Jul 7, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Git for Windows crash when I try to use git difftool -d on a file added with -N option #2677

Git for Windows crash when I try to use git difftool -d on a file added with -N option #2677

sachinrk commented Jun 9, 2020 •

edited

Loading

dscho commented Jun 9, 2020

sachinrk commented Jun 9, 2020

dscho commented Jun 10, 2020

dscho commented Jul 2, 2020

sachinrk commented Jul 2, 2020 via email

dscho commented Jul 8, 2020

sachinrk commented Jul 8, 2020 via email

dscho commented Jul 8, 2020

sachinrk commented Jul 8, 2020

sachinrk commented Jul 20, 2020 •

edited

Loading

dscho commented Aug 10, 2020

Git for Windows crash when I try to use git difftool -d on a file added with -N option #2677

Git for Windows crash when I try to use git difftool -d on a file added with -N option #2677

Comments

sachinrk commented Jun 9, 2020 • edited Loading

Setup

Details

dscho commented Jun 9, 2020

sachinrk commented Jun 9, 2020

dscho commented Jun 10, 2020

dscho commented Jul 2, 2020

sachinrk commented Jul 2, 2020 via email

dscho commented Jul 8, 2020

sachinrk commented Jul 8, 2020 via email

dscho commented Jul 8, 2020

sachinrk commented Jul 8, 2020

sachinrk commented Jul 20, 2020 • edited Loading

dscho commented Aug 10, 2020

sachinrk commented Jun 9, 2020 •

edited

Loading

sachinrk commented Jul 20, 2020 •

edited

Loading