From: "Christopher Nelson" To: Subject: RELOC_XXXXX Date: Sat, 17 Apr 1999 21:44:15 -0600 Message-ID: <01be894d$bb12a140$LocalHost@thendren> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0026_01BE891B.70783140" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.71.1712.3 X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3 Reply-To: djgpp AT delorie DOT com This is a multi-part message in MIME format. ------=_NextPart_000_0026_01BE891B.70783140 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Thanks for all the help you've given me up to now, but I have discovered = another problem that I can't find the solution too, and is related to = the same area. =20 The DOCS say: RELOC_ADDR32 To do this relocation, you must perform the following steps:=20 a.. Get the address of the symbol referred to.=20 b.. Add the value currently stored in the location being adjusted.=20 c.. Store the value back into the location being adjusted.=20 =20 This is fine if the object contains only one section, but when it = contains multiple sections, it leaves a little bit of specification. According to the paragraph above in the instructions, "If the symbol is = in the same object, the symbol table will have entries that refer to the = sections themselves (always there and always private) that will be = referred to. When you relocate the section itself, these symbols will = reflect its new location." The symbol I'm trying to adjust in this instance is a global integer = located in a section called .xptdata. This is the code I use to readjust a RELOC_ADDR32: (the 'a' below is a loop variable that goes from 0 to = file_header.f_nscns-1) unsigned long *sc; unsigned long offset, new_base; unsigned long orig; sc =3D (unsigned long *)sections[a].data; offset =3D r->r_vaddr; // - sections[a].header.s_paddr; new_base =3D (unsigned = long)sections[sym_table[r->r_symndx].e_scnum-1].data;=20 *(sc+offset) +=3D (unsigned long)new_base; The description above doesn't say whether I should use the base address = of the section that the variable is contained in (which sounds = reasonable) or whether it needs to be in the section that contains the = reloc entry. (or, in other words, if the value i add in is the base of = the section where the variable is found, or the base of the section that = the relocation entry is part of.) The problem is, this doesn't seem to work. The relocated code executes = fine, but the address for the variable is 0x40, which is definitely not = right. This is the offset that I get in the r->r_vaddr member too. What am I overlooking? -=3D{C}=3D- ------=_NextPart_000_0026_01BE891B.70783140 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Thanks for all the help you've given = me up to=20 now, but I have discovered another problem that I can't find the = solution too,=20 and is related to the same area.

The DOCS say:

RELOC_ADDR32

To do this relocation, you must perform the following steps:=20

Get the address of the symbol referred to.=20
Add the value currently stored in the location being adjusted.=20
Store the value back into the location being adjusted. =

This is fine if the object contains only one section, but when it = contains=20 multiple sections, it leaves a little bit of specification.

According to the paragraph above in the instructions, "If the = symbol=20 is in the same object, the symbol table will have entries that refer to = the=20 sections themselves (always there and always private) that will be = referred to.=20 When you relocate the section itself, these symbols will reflect its new = location."

The symbol I'm trying to adjust in this instance is a global = integer=20 located in a section called .xptdata.

This is the code I use to readjust a RELOC_ADDR32:

(the 'a' below is a loop variable that goes from 0 to=20 file_header.f_nscns-1)

unsigned long *sc;
unsigned long offset, new_base;
unsigned = long=20 orig;

sc =3D (unsigned long=20 *)sections[a].data;
offset =3D r->r_vaddr; // -=20 sections[a].header.s_paddr;
new_base =3D (unsigned=20 long)sections[sym_table[r->r_symndx].e_scnum-1].data;

*(sc+offset) +=3D (unsigned long)new_base;

The description above doesn't say whether I should use the base = address of=20 the section that the variable is contained in (which sounds reasonable) = or=20 whether it needs to be in the section that contains the reloc = entry. (or,=20 in other words, if the value i add in is the base of the section where = the=20 variable is found, or the base of the section that the relocation entry = is part=20 of.)

The problem is, this doesn't seem to work. The relocated code = executes fine, but the address for the variable is 0x40, which is = definitely not=20 right. This is the offset that I get in the r->r_vaddr member=20 too.

What am I overlooking?

=20 -=3D{C}=3D-

------=_NextPart_000_0026_01BE891B.70783140--