Re: [hackers] [quark][PATCH] Fix buffer over-read in decode()

On Tue, 2022-08-16 at 21:32 +0200, Hiltjo Posthuma wrote: > Haven't tested the patch and not sure it is correct, but if so then > isxdigit needs a cast using (unsigned char). Thanks, I don't have a lot of practical C experience. Or experience with submitting code through email, I seem to have

Re: [hackers] [quark][PATCH] Fix buffer over-read in decode()

On Tue, Aug 16, 2022 at 08:58:37PM +, HushBugger wrote: > Thanks, I don't have a lot of practical C experience. The reason for the cast is because is a poorly designed library where the caller needs to ensure that the arg is representable as an `unsigned char` (i.e 0 .. UCHAR_MAX) or as

Re: [hackers] [sbase] [PATCH] Use ar(1)'s s-flag instead of invoking ranlib(1)

On Mon, 1 Aug 2022 11:09:16 +0200 "Roberto E. Vargas Caballero" wrote: Dear Roberto, > Because then you will support only the last systems. If you keep > the ranlib you will support systems that support all versions of > the standard. Again, if you find a system without ranlib then > we can

[hackers] [libgrapheme] Separate CC into CC and BUILD_CC, add option to disable ldconfig(1)-call || Laslo Hunhold

commit 68391fc580b9a855271be57038d69cd2ce6ca0de Author: Laslo Hunhold AuthorDate: Tue Aug 16 18:56:21 2022 +0200 Commit: Laslo Hunhold CommitDate: Tue Aug 16 18:56:21 2022 +0200 Separate CC into CC and BUILD_CC, add option to disable ldconfig(1)-call The

Re: [hackers][quark][patch] pre-compression

On Mon, 8 Aug 2022 07:57:51 -0400 fo...@dnmx.org wrote: Dear fossy, > I'll try to do it by myself, but I don't promise anything.. It seems > like the resp files were moved to esnprintf or something. > > Sorry for no original e-mail text, DNMX is broken :d given quark's new structure

Re: [hackers][quark][patch] pre-compression

On Tue, 16 Aug 2022 11:19:55 -0400 fo...@dnmx.org wrote: Dear fossy, > Ah.. so very complicated, huh? Oh, well.. it's not a low-hanging-fruit by any means, yes. > Hey, so.. I have a question.. how come that Suckless' web-site isn't > hosted using Quark? If I remember correctly - it's Nginx? >

Re: [hackers] [libgrapheme] [PATCH] Remove dead file `src/util.c'

On Wed, 10 Aug 2022 15:03:06 + Tom Schwindl wrote: Dear Tom, > Since commit 072bb271868a3583da1f ("Introduce mostly branchless > character break detection") removed the code from the file, it no > longer serves a purpose. while it was true, it now is used again. :) It often makes more

Re: [hackers] [libgrapheme] Rename reallocarray() to reallocate_array() to prevent mangling || Laslo Hunhold

On Mon, 1 Aug 2022 21:53:45 +0600 NRK wrote: Dear NRK, > Given that this no longer shadows a libc/conventional function, I'd go > one step further and move the `fprintf + exit` check inside > reallocate_array() so the calling code doesn't need to worry about > null returns. thanks for your

Re: [hackers] [libgrapheme] Use (size_t)(-1) instead of SIZE_MAX and fix style || Laslo Hunhold

On Sun, 31 Jul 2022 12:18:22 +0200 Mattias Andrée wrote: Dear Mattias, > Why wouldn't SIZE_MAX be the maximum of size_t? you're totally right and I changed it. Thanks! With best regards Laslo

[hackers] [libgrapheme] Use SIZE_MAX instead of (size_t)(-1) || Laslo Hunhold

commit 5ef98fdec958130a92aa40006ddd9daba36b4d12 Author: Laslo Hunhold AuthorDate: Tue Aug 16 16:51:17 2022 +0200 Commit: Laslo Hunhold CommitDate: Tue Aug 16 16:51:17 2022 +0200 Use SIZE_MAX instead of (size_t)(-1) Thanks Mattias for pointing this out!

[hackers] [libgrapheme] Revert "Use ar(1)'s s-flag instead of invoking ranlib(1)" || Laslo Hunhold

commit 3697387a0a5fcb53df0094d988496f7a681ae54c Author: Laslo Hunhold AuthorDate: Tue Aug 16 17:43:07 2022 +0200 Commit: Laslo Hunhold CommitDate: Tue Aug 16 17:43:07 2022 +0200 Revert "Use ar(1)'s s-flag instead of invoking ranlib(1)" After further discussion with Roberto

Re: [hackers] [quark][PATCH] Fix buffer over-read in decode()

On Tue, Aug 16, 2022 at 05:42:50PM +, HushBugger wrote: > The format specifier for parsing percent-formatted characters uses > a maximum number of digits, not an exact number of digits. > > If the hex number has only one digit this will skip a character, > potentially pointing past the

[hackers] [libgrapheme] Move get_codepoint_*()-util-functions to src/util.c || Laslo Hunhold

commit aa5dda2687c4907d6a47e57b1d7973b8f9d158ae Author: Laslo Hunhold AuthorDate: Tue Aug 16 16:25:31 2022 +0200 Commit: Laslo Hunhold CommitDate: Tue Aug 16 16:25:31 2022 +0200 Move get_codepoint_*()-util-functions to src/util.c Signed-off-by: Laslo Hunhold diff --git

Re: [hackers] [sbase] [PATCH] Use ar(1)'s s-flag instead of invoking ranlib(1)

Hi, On Tue Aug 16, 2022 at 5:47 PM CEST, Laslo Hunhold wrote: > On Mon, 1 Aug 2022 11:09:16 +0200 > "Roberto E. Vargas Caballero" wrote: > > Dear Roberto, > > > Because then you will support only the last systems. If you keep > > the ranlib you will support systems that support all versions of >

[hackers] [quark][PATCH] Fix buffer over-read in decode()

The format specifier for parsing percent-formatted characters uses a maximum number of digits, not an exact number of digits. If the hex number has only one digit this will skip a character, potentially pointing past the terminating null byte. --- http.c | 3 ++- 1 file changed, 2