In T6606#173044, @gniibe wrote:

More care is needed to be perfect; There are places in GnuPG where assuan_sock_connect may be used before syscall clamp set up (after the first assuan_sock_bind failure).

Dear Werner, have you had any toughts about this ?

Currently, Kleopatra cannot do anything about this. get_passphrase in protect-tool.c asks those questions and doesn't support a way to give the user more context (e.g. by providing the file name). Once gpg-agent allows giving context, Kleopatra can add for example the file name to the data to import.

Applied to master.

yes, one down, two to go...

Meanwhile the AppImage (same binaries as the current Gpg4win version) can be found here among the binary releases: https://gnupg.org/download/index.html

I realized again how bad the current implementation is last week when Alexander managed to send a mail to me encrypted with a completely unrelated key.
a) It was not clear to him that he encrypted to a totally different key because it only displays the keyid
b) He somehow managed to store that key for me in the addressbook
c) He again selected something like "always encrypt to this user" in the dialog without realizing the consequences. This created a contact for me in his personal address book (invisible to him because he said he does not use the addressbook and in there all sources were unselcted) which had the wrong keyid (again only the fingerprint there) and the setting "always encrypt to this user"

Applied the changes for libassuan T6487 into gniibe/t6606.

Pushed the change in gniibe/t6606 branch.

Another approach would be:

• Use assuan_sock_accept which has consistent API with gnupg_fd_t

On 64-bit Windows, the situation now is:

Use of FD2INT for the first argument of select is semantically not good. It's the number of file descriptor. When we use FD2INT here, the type is converted to 64-bit integer, then implicitly demoted to 32-bit integer. We need new macro, say, FD2NUM to convert FD into 32-bit integer.
<--- done in: rGea1935252e28: commond: Introduce FD2NUM to express conversion to number of fds.

Here is a test program for 64-bit Windows to see how cast works:

#include <stdint.h>
#include <stdio.h>

It turned out that my pinentry reported "fully canceled" on Cancel (see T6491: Pinentry-Qt: Password prompt for each subkey if password change is cancelled) which made gpg output nothing.

Tested and works now for me as expected. Thanks.

The original reporter mentioned that this only occurs when called from kleo. But let me recheck.

Also done for 2.2.

This has long been implemented due to the backport of the P12 parser and the recent rewrite of it.

gpg --export-secret-subkeys --armor 704769B8D5C15319A27C74BBB47052506607DA6E confirms that gpg 2.4.1-beta21 outputs nothing if the password entry is canceled.

Of course, it's about right clicking the encryption subkey. That's what I tested. Anyway, cancel wasn't handled properly. Now it is.

In T5755#172514, @ikloecker wrote:

I cannot reproduce the problem with Cancel. When I try this, I get the error "The result of the export is empty." and nothing is written to disk. I'm using GnuPG 2.4.

Anyway, handling of cancel was indeed missing.

The expiry checker checks for expiry. It doesn't and shouldn't do anything else.

I cannot reproduce the problem with Cancel. When I try this, I get the error "The result of the export is empty." and nothing is written to disk. I'm using GnuPG 2.4.

Another request for this would be that the for expired keys a --locate-key might be triggered. GpgOL currently does this in internal logic and this causes GnuPG to refetch the key e.g. from WKD if the key came originally from WKD. https://bugs.kde.org/show_bug.cgi?id=471911 I am not sure if the expiry checker already does this, but someone pointed me to the KDE bug and I will point back here because it makes little sense to fix this in the kmail resolver when we want to replace it.

This has a serious usability issue. If you cancel the password entry when exporting it reports success and creates an apparently valid secret key file but without the subkey you intended to export. So worst case the user thinks he has a backup but instead has no backup :/

The case in check_special_filename is fixed. So, there is no cases in GnuPG where the value of out of range is silently converted to wrong value.

Remaining places are:

• common/sysutil.c debug output to format HANDLE --> T6597
• common/sysutil.c iobuf_get_fd and its use cases --> T6580
• agent, kbx, scdaemon, dirmngr: socket and FD2INT/INT2FD --> T6598

works

Except a case, all use cases of translate_sys2libc_fd_int is with a result of integer from command line argument.

Changes are pushed.

We need to keep the gpgtar part of commit in rG2756147e392c: gpg,sm,tools: Use string for option --*-fd..

The changes are intrusive to other implementations (POSIX and Windows 32-bit).
So, I revert the changes of replacing translate_sys2libc_fd_int.

I don't argue about the technical necessity for the change. I agree the fact it works (without such changes).

Just to clarify this change for readers not accustomed to Windows internals: This function was used to translate the file descriptor as passed to gpg (which is a HANDLE) to the libc file descriptor as used by stdio. Obviously we won't anymore work with stdio file descriptors in the future but use the Windows32 API (ReadFile et al). libc fds 0,1,2 are handled in a special way on Windows.

Fixed in master.

See for T6545 for a new request to support IDP.

We had one request to support this back in 2017 but it was closed because the respective CA stopped using this extension. See T2039.

The use cases are:

• oPassphraseFD for gpgsm, gpg
• oStatusFD for gpg-auth, gpg-wks-client, gpg-card, gpg-pair-tool, gpgtar, gpgconf, gpgsm, gpg, gpgv
• oLoggerFD for gpgsm, gpg, gpgv
• oAttributeFD for gpg
• oCommandFD for gpg
• oOverrideSessionKeyFD for gpg

rGb1ecc8353ae3 is just what I meant, so that we can recommend such an option in the future as a workaround until a new update becomes available which supports such an extension.

Nah, the description for that extension is pretty strict and I won't feel comfortable to just ignore it. BTW there is also T6398 (nameConstraints) which needs support. But for debugging a ignore extension makes sense.

For support reasons I would say that it might make sense to also ignore the extensions from "ignore-cert-extension" when checking CRLs?

I tested this with OpenPGP and 2.4.3-beta19 on Windows. Worked nicely.

And of course we also need to adjust GPGME

We also need PROGRESS lines in gpgsm.

I'm reopening this. Its probably not a regression but I was sure that we had progress for large files fixed in the past.

Yeah no progress for files larger then 32 bit o.O... But this used to work 😭

On 64 bit linux this works btw. so I think it comes down to the difference between 32 bit off_t and 64 bit off_t

Yeah, its the ugly off_t again. I am just testing how this works with single files above that threshold we worked quite a bit on this back in the days https://dev.gnupg.org/T2368

Yeah, probably a Windows/MinGW 32-bit problem. GpgME::Data does

off_t size = seek(0, SEEK_END);
seek(0, SEEK_SET);
std::string sizestr = std::to_string(size);
// Ignore errors as this is optional
gpgme_data_set_flag(d->data, "size-hint", sizestr.c_str());

Probably some issue with large files / integer overflow. I am testing on Windows with 32 bit.