Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Romain Naour <romain.naour@gmail.com>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH] package/lshw: fix musl build
Date: Thu, 4 Aug 2016 22:34:35 +0200	[thread overview]
Message-ID: <424edffc-502e-c9f5-d5cb-b951b0c62832@gmail.com> (raw)
In-Reply-To: <20160804200947.64225ece@free-electrons.com>

Hi Thomas,

Le 04/08/2016 ? 20:09, Thomas Petazzoni a ?crit :
> Hello,
> 
> On Thu, 4 Aug 2016 19:09:25 +0200, Romain Naour wrote:
> 
>> Actually, I get an error on basename from sysfs.cc
>>
>> sysfs.cc: In function ?std::string sysfs_getbustype(const string&)?:
>> sysfs.cc:103:41: error: invalid conversion from ?const char*? to ?char*?
>> [-fpermissive]
>>        "/devices/" + basename(path.c_str());
>>
>> The cast in dasd is not necessary.
> 
> Are you sure it's musl related, and not gcc version related? A
> -fpermissive error typically indicates that the compiler has become
> stricter than it used to be.

It seems related to musl. lshw build fine with a toolchain with gcc 5.

I'm not a glibc expert but it seems that glibc accept a char * or const char *
as filename argument, see string.h from glibc:

#  ifdef __CORRECT_ISO_CPP_STRING_H_PROTO
extern "C++" char *basename (char *__filename)
     __THROW __asm ("basename") __nonnull ((1));
extern "C++" const char *basename (const char *__filename)
     __THROW __asm ("basename") __nonnull ((1));
#  else
extern char *basename (const char *__filename) __THROW __nonnull ((1));
#  endif
# endif

The musl version doesn't support the const char * prototype.

Also, the lshw code expect the GNU version of basename(). But this version is
not available from string.h when C++ is used.

See string.h from musl:
#ifndef __cplusplus
char *basename();
#endif

That's the issue reported by autobuilders, so the only way is to use the POSIX
basename() if we want to use lshw with musl.

Upstream fixed a similar issue in src/core/pci.cc using the POSIX basename() [1]
and a cast from const char * to char * to avoid invalid conversion.

Thought ?

Best regards,
Romain

[1] https://github.com/lyonel/lshw/commit/cd690bff1516b40fecd5ec4a7f6619e5bffc3cf0

> 
> Thomas
> 

  reply	other threads:[~2016-08-04 20:34 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-08-04 16:24 [Buildroot] [PATCH] package/lshw: fix musl build Romain Naour
2016-08-04 16:37 ` Thomas Petazzoni
2016-08-04 17:09   ` Romain Naour
2016-08-04 18:09     ` Thomas Petazzoni
2016-08-04 20:34       ` Romain Naour [this message]
2016-08-05  4:25   ` Khem Raj
2016-08-12 22:41     ` Yann E. MORIN
2016-08-13  2:32       ` Khem Raj

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=424edffc-502e-c9f5-d5cb-b951b0c62832@gmail.com \
    --to=romain.naour@gmail.com \
    --cc=buildroot@busybox.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox