X-Authentication-Warning: delorie.com: mail set sender to djgpp-workers-bounces using -f
X-Recipient: djgpp-workers AT delorie DOT com
Message-ID: <590CB802.6000903@gmx.de>
Date: Fri, 05 May 2017 19:36:02 +0200
From: "Juan Manuel Guerrero (juan DOT guerrero AT gmx DOT de) [via djgpp-workers AT delorie DOT com]" <djgpp-workers AT delorie DOT com>
User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; de; rv:1.9.2.13) Gecko/20101206 SUSE/3.1.7 Thunderbird/3.1.7
MIME-Version: 1.0
To: djgpp-workers AT delorie DOT com
Subject: Re: gcc-7.1.0
References: <6953e282-24aa-2b4e-2141-385d8b443111 AT iki DOT fi> <590A6154 DOT 8010308 AT gmx DOT de> <83a86shf9d DOT fsf AT gnu DOT org> <590BA4E2 DOT 4000406 AT gmx DOT de> <49ae1668-e108-5e0b-350c-533d29574dba AT iki DOT fi>
In-Reply-To: <49ae1668-e108-5e0b-350c-533d29574dba@iki.fi>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-Provags-ID: V03:K0:5hyKxP6G1NcjUPjfwu53P9WKYErGrj35wC3zbn0j2ODqIGigh9o
 jQg43qqIjsBnxpz/ZxDif6iGonCtTqnM4nctvojFp3b96J/h1rIyl8YIfsaaf50ltkrMOAq
 rm2I7p06F0DH/uU8szQg4kIsIQcHY+Yzywr/0xYmK7zlh8ABrtuPFd8H8wFp3EtmnFabSIo
 BHczlFmAgXZPDgZ0C2f/A==
X-UI-Out-Filterresults: notjunk:1;V01:K0:q88L6/Min4M=:anguLP0V3WrABt9mJ8HO4w
 XXLX1g0eo1mCKgVZDtA6gS7Yx5Qqdks+cNo4JDMUyAvJ0UIyM/Cbouj1eNfBlKyXC5cY8xUzw
 0xPbOwbp4eB9tKxLCLQ6b4g53F13hEMGnEEIEDThFoap2khwLPZ410S2qUkmrQ18zdyhUfo4Y
 OXJR6JfhPrEg/Gtantg43eXQsRy2VN2K+C/LRexQmjIoBvIq0q+fbi8GRRw2f7gUkBkWpZPCX
 1EkbxrlP4ku3qGcRdKSoSsWjkuU6mJrHOvkwLCM9CJkfbGT1KhWDDs8fCszSmytxJgN+4sz44
 +KqmuKp/D7ReMIE88o6t2C3BMNt+1crnOCs/gizFgRCW9UVFBaeeldioLkieRZuDhLiAMk4wW
 t51rANSxxxe4wkiju0akJYWSmP5i+7eh5E1IJ7xibG37TH8+dX/ogBRPaw2sMsrx8pZbFQpem
 Df/VgQAlXxTxCt3N2KDJMEy2T+jjfDD7s1XC7qkbLpLgq36fe6VoZnjb6hhZAdBDu2ytRbxip
 TuVi9PWLOJ1rfd0V0NK/4t8yrP6IzTUtTt1QwBW8aXHtpQqnE8k/OJSWxnn+BsWu1ctas0IPR
 dpCvA0EvxsaSPOPGTyK1z5+Z8lnlDckqHYaI+zCtS9Mu5oHntL/owq75h1OlXWxo5rYvddIPS
 6mAMg1MbTVW4gY5pVuLYVKlwMtNQrxnXT+x8j+hlUvQ6h0BxSeqoeWeKja6U55DIl+GmA5bAo
 2uoq/eNoKJjyXs3qZnqNY3Ww0Aes5P7j2oyp55dBQU3kBy3Z58BdeMpPBIyerVF029t8AOH/J
 RIf8Me+
Reply-To: djgpp-workers AT delorie DOT com

Am 05.05.2017 06:12, schrieb Andris Pavenis (andris DOT pavenis AT iki DOT fi) [via djgpp-workers AT delorie DOT com]:
> On 05/05/2017 01:02 AM, Juan Manuel Guerrero (juan DOT guerrero AT gmx DOT de) [via djgpp-workers AT delorie DOT com] wrote:
>> Am 04.05.2017 16:23, schrieb Eli Zaretskii (eliz AT gnu DOT org) [via djgpp-workers AT delorie DOT com]:
>>>> Date: Thu, 04 May 2017 01:01:40 +0200
>>>> From: "Juan Manuel Guerrero (juan DOT guerrero AT gmx DOT de) [via djgpp-workers AT delorie DOT com]"<djgpp-workers AT delorie DOT com>
>> [snip]
>>>
>>> It's probably best to tweak the code to make sure the buffer overruns
>>> cannot happen.
>>
>> I have tried to fix the buffer overruns by recomputing the length correctly
>> so the compiler stops complaining. Also some minor warnings in the zoneinfo
>> code has been fixed but they do not stop compilation. I have added "break"
>> in switch statements where they are missed. Neitherless there are some
>> warnings due to intentional fall throughs that may be suppressed by excluding
>> the corresponding compiler warning settings [-Wimplicit-fallthrough=].
>>
>> As usual suggestions, objections, comments are welcome.
>>
>>
>> Regards,
>> Juan M. Guerrero
>>
>
> I guess better way is to add explicit casts to unsigned short. Attached patch seems to be sufficient.
> Tested however build only using Linux to DJGPP cross-compiler.
>
> Andris
>
>


Index: dosexec.c
===================================================================
RCS file: /cvs/djgpp/djgpp/src/libc/dos/process/dosexec.c,v
retrieving revision 1.27
diff -p -r1.27 dosexec.c
*** dosexec.c	2 May 2015 07:32:23 -0000	1.27
--- dosexec.c	5 May 2017 04:04:28 -0000
*************** direct_exec_tail_1 (const char *program,
*** 406,412 ****
     if ((initial_tbuf_selector != tbuf_selector) && proxy_off)
     {
       char temp[65], *s, t2[5];
!     sprintf (t2, "%04lX", tbuf_beg>>4);
       dosmemget (tbuf_beg+proxy_off, 64, temp);
       temp[64] = 0;
       s = strchr(temp,'\r');
--- 406,412 ----
     if ((initial_tbuf_selector != tbuf_selector) && proxy_off)
     {
       char temp[65], *s, t2[5];
!     sprintf (t2, "%04X", (unsigned short)(tbuf_beg>>4));                      <--   OK with me
       dosmemget (tbuf_beg+proxy_off, 64, temp);
       temp[64] = 0;
       s = strchr(temp,'\r');
*************** static int go32_exec(const char *program
*** 1107,1115 ****
     proxy_cmdline = (char *)alloca (34);

     sprintf(proxy_cmdline, "%s=%04x %04x %04x %04x %04x",
!     __PROXY, argc,
!    (unsigned)(tbuf_beg >> 4), rm_off & 0xffff,
!    (unsigned)(tbuf_beg >> 4), si_off & 0xffff);
     if (!found_si)
       proxy_cmdline[22] = 0; /* remove stubinfo information */

--- 1107,1115 ----
     proxy_cmdline = (char *)alloca (34);

     sprintf(proxy_cmdline, "%s=%04x %04x %04x %04x %04x",
!     __PROXY, (unsigned short)argc,                                            <--   not OK with me
!     (unsigned short)(tbuf_beg >> 4), (unsigned short)(rm_off & 0xffff),       <--   OK with me
!     (unsigned short)(tbuf_beg >> 4), (unsigned short)(si_off & 0xffff));      <--   OK with me
     if (!found_si)
       proxy_cmdline[22] = 0; /* remove stubinfo information */


The changes are OK for me but I do not understand at all why you are assuming that an allocation
of 34 bytes for proxy_cmdline shall be sufficient.  It shall contain  __PROXY and this string
is supposed to have a size of __PROXY_LEN bytes.  Or am I missing something?  That the compiler
does not complain may not imply that it is capable to really calculate the correct length.
Neither less, please feel free to fix it as you prefer.  I agree that the implemented fix shall
make minimal changes.

Regards,
Juan M. Guerrero